PDF hosted at the Radboud Repository of the Radboud University
Nijmegen
The following full text is a publisher's version.
For additional information about this publication click this link.
http://hdl.handle.net/2066/25448
Please be advised that this information was generated on 2017-06-14 and may be subject to
change.
Copyright © E R S Journals Ltd 1997
European Respiratory Journal
IS S N 0 9 0 3 - 1936
Eur Respir J 1997; 10: 2 6 6 2 - 2 6 8 9
Printed in U K - all rights reserved
ERS TASK FORCE
Clinical exercise testing with reference to lung diseases:
indications, standardization and interpretation strategies
ERS Task Force on Standardization of Clinical Exercise Testing
Members o f the Task Force: J. Roca and B.J. W hipp (co-chairmen), A.G.N. Agusti, S.D. Anderson,
R. C asaburi, J.E. Cotes, C.F. Donner, M. Estenne, H. Folgering, T.W. Higenbottam, K.J. Killian,
P. Paiange, A. Patessio, C. Prefaut, R. Sergysels, P.D. Wagner and I. W eism an
CONTENTS
Responses to exercise in lung disease....
2663
Pulmonary gas exchange...................... .................. 2663
Ventilation ............................................................ . 2663
Arterial blood gases ............................................... 2665
Cardiovascular response ......*................................. 2665
Limits to exercise.................................................. 2666
Indications.......................................................... 2666
Assessing exercise tolerance and potential limiting factors
, 2667
Assessment of impairment in chronic lung diseases. 2667
Preoperative assessment ......................................... . 2667
Rehabilitation programmes: patient assessment and
exercise prescription.................... ........................... .. 2668
Assessment of impairment-disability....................... .. 2668
Lung, heart-lung transplantation................................ 2668
Protocols.................................................................. 2668
Standard clinical exercise protocol: incremental test.. 2668
Other tests:
Constant work rate exercise testing........................ 2669
Assessment of exercise-induced bronchoconstriction 2670
Equipment and quality-control programme.......... 2671
Choice of ergometers.............................................. 2671
Gas exchange measurements................................... 2672
The lactate threshold.............................................. 2673
Exercise testing measurement systems.................... 2673
Quality control programme..................................... 2615
Personnel and testing procedures......................... 2676
Personnel qualifications.......................................... 2676
Patient preparation.................................................. 2676
Safety precautions................................................... 2677
Interpretative strategies........................................ 2678
General guidelines.................................................. 2678
Reference values..................................................... 2680
Appendix............................................................ 2681
Cardiopulmonary exercise testing (CPET) is a unique
tool to assess the limits and mechanisms of exercise
tolerance. It also provides indices of the functional
reserves of the organ systems involved in the exercise
response, with inferences for system limitation at peak
exercise. Moreover, CPET is useful for establishing the
profiles and adequacy of the responses of the systems
at submaximal exercise. The present document is essen­
tially focused on clinical problems commonly faced in
the study of patients with pulmonary diseases. Physiologi­
cal changes of the respiratory system during exercise,
however, should only be considered as part of a co-ordinated sequence of oxygen and carbon dioxide transfer
processes between the atmosphere and the mitochon­
dria to meet the increased energy demand of the skele­
tal muscle. Consequently, even in the analysis of patients
with well-identified pulmonary disease, an integrative
approach to CPET [1, 2] is required.
CPET is an area of growing interest in pulmonary
medicine for three major reasons: 1) its large potential
clinical applicability (see section on Indications); 2) the
essentially noninvasive nature of the testing; and 3) pro­
vision of information that cannot be obtained through
conventional lung function testing performed at rest [39]. During the past few years, two factors have con­
tributed to the current level of interest in CPET in pul­
monary medicine. First, substantial progress has been
made in clarifying fundamental concepts of exercise
physiology (e.g., factors limiting maximal oxygen up­
take, lactate threshold) which have historically been
the focus of controversy. Secondly, major technologi­
cal improvements have facilitated data collection, sub­
ject monitoring during the test and subsequent formatting
and analysis of the results. Nowadays, CPET can be
considered a primary test in the pulmonary function
laboratory.
The present European Respiratory Society (ERS) pos­
ition document reflects the views on the topic shared
by the members of the Task Force. One of the self-impos­
ed goals of the group was to produce a relatively readerfriendly document that combined a rigorous conceptual
approach with practical utility for CPET in a clinical
setting. The document can be either read as a whole,
or the first section (Responses to exercise in lung dise­
ase) can be used alone, as a frame of reference to clarify
specific points of the document. Definitions, abbreviations,
....................................................................... . .
.
....
. .
This position document of the European Respiratory Society was officially adopted by the ERS Executive Committee on March 1, 1997.
The Task Force on Standardization of Clinical Exercise Testing was endorsed by the Clinical Physiology Assembly and the Clinical Assembly
of the ERS.
Correspondence: J. Roca, Servei de Pneumologia, Hospital Clinic, Barcelona» Spain; or B J. Whipp, Dept of Physiology, St. George's Hospital,
London, UK.
CLINICAL EXERCISE TESTING
information content and units of the recommended vari­
ables are presented in the Appendix. Finally, as a group,
the Task Force acknowledges that the introduction of
SI units has represented a valuable effort not only to
provide a coherent system of units, but also to ensure
that units are uniform in concept and style. Unfortuna­
tely, the use of SI is still insufficiently widespread in
clinical exercise testing. In order to avoid nomenclature
that can be unfamiliar to many readers, we decided to
use traditional units in the text but also provide the
equivalences between the two systems of units in the
Appendix.
Responses to exercise in lung disease
Exercise intolerance results when a subject is unable
to sustain a required work rate (WR) sufficiently long
for the successful completion of the task. The cause,
most commonly, is an oxygen demand that exceeds the
maximal oxygen conductance capability of the oxygen
transport chain; the consequence is a perception of limb
fatigue, breathlessness or even, in some conditions, frank
pain. Subjects with lung disease often experience ex­
ercise intolerance at extremely low work rates. There
are many kinds of lung disease, however, and in any
one patient the structural and functional severity of the
disease may range from the barely discernible to the
very severe. As a result, responses to exercise in patients
with lung disease do not show the tight stereotypical
pattern of normal subjects.
The appropriateness of the integrated systemic respon­
ses are perhaps best studied utilizing incremental ex­
ercise testing either as a ramp or small W R increments
each of short duration. This provides a convenient means
of: 1) determining whether the magnitude and pattern
of response of particular variables is normal with respect
to other variables or to WR; 2) establishing a subject's
limiting or maximum attainable value for physiological
variables of interest; and 3) establishing exercise inten­
sity domains, such as the transition between moderate
and heavy intensity exercise. It is important to recog­
nize, in this context, the difference between submaximal
and maximal exercise levels. In submaximal exercise,
the components of the oxygen transport pathway can
provide adequate oxygen and carbon dioxide fluxes
between the air and the mitochondria. Mitochondrial
oxidative capacity has not been reached, symptoms are
usually tolerable and muscle fatigue has not occurred,
or at least may be insufficient to impair performance
appreciably. At maximal exercise, symptoms have cau­
sed the patient to stop exercising. At this stage, one or
more of the following possibilities exist: 1) Limits to
oxygen transport have been reached and maximal oxy­
gen consumption (V'o2,max) attained; under such condi­
tions, breathing 100% oxygen, for example, could increase
V’o2,max [10]; 2) Mitochondrial oxidative capacity has
been reached and again the subject would be conside­
red to be at Vc^max, but adding oxygen would not raise
oxygen consumption (V'02); 3) Maximal exercise has
occurred at a level that does not require maximal oxy­
gen transport or maximal oxidative capacity and here
exercise has been limited by unusually severe symp­
toms. Under these conditions a plateau in oxygen uptake
2663
(V o 2,max) has not been reached and the appropriate term
is peak rather than maximal V'02.
Pulmonary gas exchange
Oxygen uptake. The peak V'o2is typically low in patients
with lung disease [2, 5, 11]. The increment in V‘o 2 in
response to incremental or steady state exercise, how­
ever, occurs in a manner that often cannot be separa­
ted from that of normal subjects. That is, the response
is linear with a slope (A V'o2/chan.ge in work rate (AWR))
for cycle ergometry of approximately 10 mL-nrnr^W *1
(i.e. within the 95% confidence limits (95% Cl) of 8.5
and 11.5 for normal subjects) [12], Yet the oxygen cost
of breathing per unit ventilation is increased in both
chronic obstructive pulmonary disease (COPD) and in­
terstitial lung disease (ILD). In COPD, airway obstruc­
tion requires more effort to move a given volume of
air, while in ILD increased elastic recoil requires more
inspiratory muscle activity. In mild disease these addi­
tional costs can be quite small; in severe disease the
expected higher Vl02 from the extra cost of breathing
may be offset by the slower oxygen transport and uti­
lization kinetics (patients also tend to be sedentary) [13].
Carbon dioxide output (V'C02). Because the tissue capa­
citance for carbon dioxide is markedly greater than for
oxygen, pulmonary carbon dioxide output will initially
not rise as fast as pulmonary uptake of oxygen, despite
similar rates of metabolic exchange. For moderate con­
stant work-rate exercise in healthy young subjects, this
effect prolongs the 30-40 s time constant of metabolic
V'C02from the start of exercise (considered to be close­
ly equivalent to that of pulmonary oxygen uptake) to a
value of 50-60 s for pulmonary carbon dioxide ex­
change [14, 15]. This transient storage effect is exacer­
bated in patients with COPD who have areas of high
alveolar ventilation/perfusion (V'a IQ 1) ratios. Com­
monly, these regions receive as much as 50% of the
alveolar ventilation but only 5% or less of the cardiac
output [16-18]. As a result, the bulk of the V'C02comes
from relatively poorly ventilated areas of the lung. This,
in addition to any overall slowing of the pulmonary
blood flow, can retard the kinetics of carbon dioxide
elimination, which is also seen in those pulmonary vas­
cular disease (PVD) patients with high VlAIQ Xregions.
Ventilation
The appropriateness of the ventilatory response to
exercise depends not on the actual level of ventilation
achieved but rather on the extent to which it subserves
its pulmonary gas exchange and acid-base regulatory
requirements. The ventilatory response to exercise, how­
ever, is influenced not only by metabolic rate (most
closely for carbon dioxide) but also by the regulated or
"set point1’ level of arterial carbon dioxide tension (Ai,C0 2)
and the physiological dead-space fraction of the breath.
Quantitatively, these are represented by the following
relationship:
V'E = [863*V'C02]/[Pa,C02*(1- Vd /VT)]
(1)
2664
J. ROCA, B J . WHIPP
where 863 is the constant which corrects for the dif­
ferent conditions of reporting the gas volumes (for a
body temperature of 37°C) and also the transformation
of fractional concentration to partial pressure; V'E is
minute ventilation in litres per minute; Pa,C0 2 indicates
alveolar carbon dioxide tension in millimetres of mer­
cury (usually taken to be equivalent to Pa,C0 2); and, Vd/
VT is the dead space (Vd) expressed as a fraction of the
tidal volume (Vt).
Minute ventilation (V'E). Lung disease typically in­
creases both ventilation at rest and the ventilatory re­
quirements for a given level of exercise [2,19, 20]. The
high work of breathing may also lead to respiratory
muscle fatigue. This may produce limiting symptoms,
although shortness of breath cannot necessarily be equa­
ted to respiratory muscle fatigue. Resting ventilation
is usually greater than normal in patients with COPD,
ILD and pulmonary vascular disease (PVD). This is
indirectly due to the presence of V'a /Q' inequality pro­
duced by the classical pathological abnormalities of
emphysema and chronic bronchitis in COPD and by
fibrosis in ILD [21]. Thus, an abnormal level of venti­
lation is required to maintain normal Pa,co2, even at rest,
when there is V'AIQ' mismatching. This is especially so
in emphysematous patients who often show not only
the expected higher ventilation needed for eucapnia as
a result of high V'A/ Q' areas, but even higher ventila­
tion as evidenced by low Pa,C0 2 (i.e. <40 mmHg (5.3
kPa)). What drives them to hyperventilate remains
obscure. Other patients with COPD have carbon dioxide
retention even at rest. The carbon dioxide retention, of
course, does not necessarily indicate that the V'E is less
than normal, only that the normal defence of Pa,C02 has
failed. In ILD and pulmonary vascular disease (PVD)
there is often considerable hyperventilation even at rest
[22-24], with Pa,C02in the 30-35 mmHg (4.0-4.7 kPa)
range. This can occur even when Pa,0 2 is at or above
60 mmHg (8.0 kPa).
Breathing pattern. In patients with chronic lung dis­
eases, the Vt tends to be lower and the respiratory fre­
quency (/R) higher for a given level of V'E [2, 25]. A
strong linear relationship between peak exercise VT and
vital capacity (VC) has been shown both in patients
with COPD and idiopathic pulmonary fibrosis (IPF),
suggesting that the differences in peak "Vt are mainly due
to abnormal respiratory mechanics. However, determi­
ning Vr at submaximal exercise is also useful. During
exercise, normal human subjects increase/R by decreas­
ing inspiratory time (fl) fractionally less than expira­
tory time (iE). As a result, the ratio of n to total breath
duration (itot) (the inspiratory duty cycle (tlltm)) in­
creases in normal humans from 0.35-0.40 at rest to
0.50-0.55 during exercise. Patients with ILD show a
similar type of response. In contrast, patients with severe
COPD often show no increase in tl/ttot with exercise,
thus preserving greater time for expiration.
Pulmonary mechanics. Does ventilation during exer­
cise reach mechanical limits in patients with lung dis­
ease? Certainly, flow rates (inspiratory and/or expiratory)
can be shown to reach the envelope of the resting max­
imal flow volume curve, particularly during expiration
in COPD [26,27] and even in some elderly normal sub­
jects who are physically fit [28]. As this may be expect­
ed to result in abnormal respiratory sensations, this may
contribute to exercise limitation. However, unless flow
limitation is occurring through a substantial majority
of expiration rather than over a small part of the breath,
there is no capacity to increase lung volume (to advan­
tage airflow) and fk cannot be increased without com­
promising VT, it may well be that total ventilation is
not at a mechanically defined upper limit, even if a por­
tion of respiratory flow is. Patients with COPD adopt
two strategies during exercise to increase V 'E when there
is expiratory flow limitation: 1) end-expiratory lung vol­
ume (EELV) increases [26, 27], in contrast to normal
humans who show a fall in EELV during exercise [29],
or in patients with ILD, who do not change EELV sig­
nificantly during exercise [29, 30]; and 2) inspiratory
flow rate increases with decreased inspiratory time (i.e.,
allowing more time for expiration). Traditionally, "ven­
tilatory limitation" during exercise has been expressed
by comparing exercise ventilation to the resting maxi­
mal voluntary ventilation (M W ) as an estimate of ven­
tilatory capacity. However, while convenient, M W is
not an ideal yardstick of whether a limit to ventilation
has occurred during exercise. The M W manoeuvre is
performed over only 12-15 s at VT and Jk values that
usually differ from those adopted naturally during ex­
ercise, and also at different end-expiratory/inspiratory
lung volumes. Another approach to assessing whether
ventilation during exercise has attained its limiting value
has been to add carbon dioxide to the inspired gas dur­
ing exercise [17,31]. If ventilation increases in response
to the added carbon dioxide, then it can be concluded
that ventilation in the absence of carbon dioxide was
not mechanically limited. However, failure to increase
ventilation in response to added carbon dioxide or to
added external dead space, for example, could be due
either to mechanical limitation or to insufficient chemoreceptor stimulation to increase ventilation in the face
of other ventilatory constraints. A further approach has
been to unload the respiratory system, by breathing 21%
oxygen in helium rather than in air. This usually in­
creases ventilation during maximal exercise in both
COPD patients [17, 32] and normal subjects [33, 34].
However, one is now dealing with a mechanically alter­
ed respiratory system with a reduced work of breath­
ing per unit ventilation. Such a result is not necessarily
evidence that ventilation while respiring room air was
mechanically limited.
Respiratory muscle function. There are several reasons
to suggest that respiratory muscle function during exer­
cise can be a limiting factor in patients with lung dis­
ease. Most prominent among them are: 1) the load to
the respiratory muscles is increased both in COPD (air­
flow obstruction) and ILD (increased lung elastance);
2) in COPD the diaphragm is at a mechanical disad­
vantage due to dynamic hyperinflation; and 3) hypoxaemia during exercise may render respiratory muscles
prone to fatigue.
Bronchomotor tone response to exercise. The normal
bronchomotor response to exercise is a mild degree of
broncliodilation. Patients with exercise-induced asthma
CLINICAL EXERCISE TESTING
and some patients with COPD may show exercise-in­
duced bronchoconstriction, but this is most commonly
a postexercise phenomenon. Exercise testing is, by def­
inition, useful in investigating the mechanisms of ex­
ercise-induced bronchoconstriction.
Arterial blood gases
Many patients with COPD, ILD or PVD are hypoxaemic at rest breathing room air at sea level. That is,
their Pa,02 is below the lower reference limit, i,e 80
mmHg (10.6 kPa). There is, however, great variability,
with Pa,02 being within normal limits for age or as low
as 30-40 mmHg (4.0-5,3 kPa). In COPD, most of the
patients show mild-to-moderate hypoxaemia 60-70
mmHg (8 .0-9.3 kPa). Those COPD patients who retain
carbon dioxide at rest are generally more hypoxaemic,
with Pa,02 often being in the 40-60 mmHg (5.3-8.0
kPa) range. A similar broad range of Pa,02 is seen in
ILD and PVD. Pa,C02 is usually normal or slightly re­
duced in those COPD patients with Pa,02 in the 60-70
mmHg (8 .0-9.3 kPa) range. Carbon dioxide retainers,
of course, have an elevated Pa,C0 2s usually in the range
45-55 mmHg (6.0-73 kPa), but again extremes exist
with values as high as 70-80 mmHg (9.3-10.6 kPa)
occasionally seen, even in stable out-patients. In ILD,
Pa,C02 is typically 30-35 mmHg (4.0-4.7 kPa), as dis­
cussed previously. This is also true in PVD. From these
general pairings of partial pressure of oxygen (P0 2) and
partial pressure of carbon dioxide (Pco2), the alveolar
gas equation can be used to deduce the alveolar-arter­
ial oxygen difference (DA-a,0 2):
DA-a,02 = (Pi,02 - Pa,C0 2/RER +
(Pa,C02*Pl,0 2*(1 -RER)/RER)) - Pa,02
(2)
where RER is the respiratory exchange ratio (RER=
V'CO2/V02) Pl£)2 is the inspiratory oxygen pressure and
Pl,02 is the inspiratory oxygen fraction. This equation
is derived from mass balance considerations of oxygen
and carbon dioxide exchange between the air and the
alveolar gas and assumes constancy of lung nitrogen
stores. Thus, even if V'02i V'C02 and V'E are changing,
the DA-a,02 can be validly computed if there is nitro­
gen balance between inspiration and expiration. It is
important to recognize that the absolute blood-gas val­
ues and the DA-a,02 yield different, but complementary
information. The DA-a,02 reflects primarily pulmonary
defects in gas exchange caused by V'AIQ' mismatch­
ing, diffusion limitation and shunt either alone or in
combination. It can, however, be modified by changes
in cardiac output or ventilation even in the absence of
change in V'AIQ' distribution. The typical response to
exercise in COPD is a small rise in Pa,C02 and a sim­
ilar, or even greater, fall in Pa,0 2. However, Pa,02 fre­
quently does not fall, and may even increase in some
subjects. Studies using the multiple inert gas elimina­
tion technique show that V'AlQ' mismatch is usually
unaltered from that at rest in COPD, that shunts do not
develop, and that diffusion limitation also does not oc­
cur [16, 35]. This is particularly so when COPD is
severe. In milder disease, there is evidence that small
improvements in V'AIQ' relationships may occur on
2665
exercise [18]. The blood-gas changes on exercise are
mostly the consequence of how alveolar ventilation
increases compared to V'02 and V'C0 2, with secondary
effects from the fall in mixed venous Po2. However, it
is not infrequently observed that when the patient with
COPD is encouraged to maximal effort, sudden hypox­
aemia and hypercapnia can develop. In ILD the bloodgas changes with exercise are usually more typical and
substantial even at moderate effort. While Pa,C02 is gen­
erally unaffected, Pa,02 falls in almost all patients with
diffusion limitation being contributory [23]. It is remar­
kable that diffusion limitation of oxygen exchange can
occur even at a cardiac output of less than 10 L*min_I
during exercise in ILD, However, as with COPD, the
amount of V'AfQ' mismatching and shunt appear not to
change with exercise [23]. In PVD, Pa,02 usually falls
by several millimetres of mercury, Pa,C02 also falls by
perhaps 3-5 mmHg (0.40-0.67 kPa), and the DA-a,0 2,
thus, increases. As with ILD, this is found to be large­
ly due to the large fall in venous Po 2 rather than a syste­
matic change in V a/Q ’ relationships [23], and diffusion
limitation may also develop.
Cardiovascular response
It should be remembered that patients with lung dis­
ease are subject to the same general cardiovascular dis­
orders as anyone else. Those with COPD in particular
share risk factors for ischaemic heart disease due to ciga­
rette smoking, age and inactivity, and in addition they
are generally hypoxaemic. However, throughout the fol­
lowing discussion, ischaemic heart disease is assumed
not to be present. Also, it is well known that eventual­
ly many patients with either COPD or ILD will devel­
op cor pulmonale with right heart failure from their lung
disease. This will also happen in PVD. While the pul­
monary circulation shows evidence of abnormality well
before frank right heart failure develops, the following
discussion is limited to those patients who have not yet
reached the stage of clinical right heart failure. Typically
cardiac output increases per unit increase in metabolic
rate even in patients with very severe lung disease [16,
36], except in some with PVD. This is true even if car­
diac output is somewhat reduced at rest. However, at
peak exercise in these forms of lung disease, cardiac
output is about 50% (less in PVD) of what a normal
older subject could achieve at peak exercise [18, 36].
There are two possible explanations. First, the control
of cardiac output during exercise in lung disease may
remain so tight that, despite the capacity for a higher
cardiac output, it remains regulated to match the level
of V'02 achieved. The second is that despite absence of
overt heart failure, cardiac function is indeed compro­
mised and a higher cardiac output could not be achiev­
ed. Pulmonary hypertension is often evident even at
rest, and usually worsens during exercise. The increase
in pressure per unit increase in cardiac output is some­
times three times greater in the patient groups than in
normal subjects. In normal subjects, although pulmonary
artery pressure normally rises during exercise, pulmon­
ary vascular resistance normally falls due to a combina­
tion of vascular recruitment and distension in the lungs.
In ILD , PVD and severe COPD on the other hand, vas­
cular resistance remains constant or may even rise [18,
2666
J. ROCA, B.J. WHIPP
23, 24]. The reason is vascular destruction or obstruc­
tion, which is known to occur in these diseases, and hypo­
xic vasoconstriction. Eventually, as the diseases progress,
the right heart will hypertrophy and ultimately fail, and
clinically significant cor pulmonale will be present.
Despite the marked increase in vascular resistance, it
is remarkable that even in advanced disease the heart
can pump in an essentially normal manner as a func­
tion of filling pressure if heart failure is not present.
Cardiac frequency
at a given
is higher than
normal in subjects with these lung diseases. This im­
plies that stroke volume must be lower* Whether this is
simply an extension of the phenomenon in health, where­
by unfit subjects have a higher
than fit persons at
the submaximal Vr0 2, is unclear. The oxygen pulse (V'o2
per heart beat) is an index that is in common current
use (see [2] pages 65-66 and 119-120 for discussion)
because: 1) of its noninvasive nature; and 2) it is quan­
titatively equivalent to the product of stroke volume and
arterial minus mixed venous oxygen difference. This
will be lower in lung disease if
is higher. However,
despite the
being higher at the same V'o2 than in
health,
at peak
is almost always less than pre­
dicted. As the V'o2-cardiac output relationship is gene­
rally normal, the arteriovenous oxygen concentration
difference will also be normal for a given V'0 2. However,
both arterial and mixed venous Po2 levels are reduced
due to the hypoxaemia of lung disease. Interestingly,
mixed venous Po2 at peak exercise in health is about
the same as in lung diseases, despite the much higher
peak V o 2 in health.
(fc)
V'o2
fc
fc
fc
fc
V]02
Limits to exercise
At levels of effort below peak or maximal, patients
with lung disease may have a total ventilation and car­
diac output increase, at a given V'02, at least as much
as in health even if respiratory muscle and cardiac work
and the patient's symptoms are greater. The physiolog­
ical basis for reduction in peak or maximal V'02 is of
major interest. The reduction in peak V o 2 in COPD
patients, for example, sometimes cannot be accounted
for simply by "ventilatory limitation"; in fact, leg fatigue
is a commonly reported end-exercise symptom in these
patients [17], Although they may not have been able to
achieve higher ventilation breathing room air because
of their lung disease, there are two additional factors
of considerable quantitative importance contributing to
the low peak V'o2. Firstly, the cardiac output at peak
exercise is about 50% of the value that a normal agematched subject would achieve at maximal effort. Whe­
ther this represents cardiac dysfunction or, as stated,
earlier manifestation of the tight regulation between
V 0 2 and cardiac output, is uncertain. However, oxygen
transport would presumably have been greater had car­
diac output reached normal maximal values and (sub­
ject to not having reached mitochondrial oxidative
limitation) the same would have been true of maximal
V'0 2. A second factor that might decrease peak V'02 is
a reduction in muscle oxygen conductance [37], result­
ing from inhomogeneity between local muscle oxygen
and blood flow or to reduced diffusional conductance,
possibly the result of myopathic capillary Tarification.
Had the muscle conductance been normal, peak V o 2
could then have been greater even at the same cardiac
output and
It is also unclear whether simple lack
of fitness or additional pathological factors are respon­
sible. O f course, if
could be improved by such the­
rapeutic gains, V'C02would rise accordingly and provide
an added burden to ventilation that might still prevent
the anticipated gain in V'O^max. Recognizing the poten­
tial capacity for factors other than ventilation to limit
exercise in lung disease is important in attempting to
correct the primary lung problem, e,g. by transplanta­
tion. For example, the poor exercise responses after
organ transplantation in cardiopulmonary diseases are
likely to be due, at least in part, to unrecognized and
untreated defects in skeletal muscle blood flow and oxy­
gen transport.
V'02
Indications
There is a range of indications for CPET, as outlined
in table 1. It is useful, for example, in the diagnosis of
a range of disease conditions, such as: 1) exercise-induced
asthma; 2) cardiac ischaemia; 3) foramen ovale paten­
cy with development of right-to-left shunt during exer­
cise [2]; and 4) McArdle's syndrome [38]. In addition,
CPET provides information on dysfunction, monitoring
and prognostic value in a wide range of conditions (table
1; [5]).
However, an adequate identification of the clinical
problem requiring study should be considered a neces­
sary prelude to CPET, as should an appropriate assess­
ment of the patient by: 1) medical history; 2) physical
examination; 3) chest radiograph; 4) pulmonary func­
tion testing; and 5) electrocardiogram (ECG). The clin­
ical problem that prompts the CPET and the specific
aims of the test {i.e. assessment of exercise tolerance,
analysis of pulmonary gas exchange during exercise,
Table 1. - indications for cardiopulmonary exercise
testing with reference to lung diseases
Assessing exercise tolerance and potential limiting factors
Identification of abnormal limitation of exercise intolerance,
and discrimination among causes of exercise tolerance
Differentiation between dyspnoea of cardiac or pulmonary ori­
gin
Evaluation of unexplained dyspnoea when initial pulmonary
lung function impairment does not provide conclusive results
Assessing impairment in chronic lung diseases
Interstitial lung diseases
Chronic obstructive pulmonary disease
Marked hypoxaemia during exercise
Occult cardiac disease
Abnormal breathing pattern during exercise
Chronic pulmonary vascular occlusion (controversial)
Cystic fibrosis
Preoperative assessment
Major abdominal surgery, especially in elderly patients
Lung cancer resectional surgery
Resectional surgery in pulmonary emphysema
Diagnosis of exercise-induced asthma
Rehabilitation programmes: patient assessment and exer­
cise prescription
Assessment of impairment-disability
Lung, heart-lung transplantation
Modified from [5].
CLINICAL EXERCISE TESTING
Identification of the
clinical problem
Clinical history
Physical examination
Pulmonary function tests
ECG
Optimize format
for presenting results
Select appropriate reference values
to establish patterns of abnormal response
Compare with characteristic patterns
of relevant diseases
Fig. 1. - Logical strategy to approach cardiopulmonary exercise
testing (CPET) in a clinical setting. ECG: electrocardiogram.
etc) determine both the type of exercise protocol to be
used and the variables to be considered in the interpre­
tation of the test (fig. 1). These are described in detail
below (see sections on exercise protocols and interpre­
tation strategies).
Assessing exercise tolerance and potential limiting fac­
tors
Exercise performance and V'02 peak cannot be val­
idly predicted from resting cardiopulmonary testing.
This is especially true in patients in whom exercise is
limited by lung function [3-9]. This is not only because
exertional dyspnoea is not well predicted from pul­
monary function, but also because many patients with
lung disease stop exercising owing to fatigue, chest pain
and leg discomfort rather than dyspnoea [3], For these
reasons and because measures of health-related quality
of life correlate better with exercise tolerance than with
either spirometry or oxygenation [9], CPET is a useful
tool in the integrated evaluation of common problems
such as unexplained dyspnoea on exertion and limita­
tion of exercise tolerance [39, 40]. Appropriate use of
CPET allows the investigator: 1) to quantify the degree
of abnormal limitation and to discriminate among cau­
ses of exercise intolerance; 2) to differentiate between
dyspnoea of cardiac or pulmonary origin when respi­
ratory and cardiac diseases coexist; and 3) to analyse
unexplained dyspnoea when initial pulmonary function
impairment does not provide conclusive results [5,40].
2667
Assessment of impairment in chronic lung diseases
Comprehensive CPET can be extremely useful in dif­
ferent phases of the decision-making process in chro­
nic lung diseases.
Interstitial lung diseases, Assessment of pulmonary gas
exchange during exercise [23, 41-43] in the clinical
evaluation of ILD is fundamental to early diagnosis,
staging of dysfunction and monitoring of therapy. More­
over, recent data in a rather small group of subjects [44]
seem to suggest that CPET might also have a progno­
stic value in idiopathic ILD.
Chronic obstructive pulmonary disease. Use of CPET
to assess the severity of dysfunction in COPD can be
valuable in the detection of situations that may prompt
therapeutic decisions to ameliorate the consequences of
the disease such as: 1) marked hypoxaemia during exer­
cise (needs for supplemental oxygen can be assessed
directly; 2) occult cardiac disease (initiate appropriate
therapeutic regime); or 3) inefficient breathing patterns
during exercise (initiate rehabilitation programme?). Use
of CPET in prospective trials may also provide inform­
ation in the evaluation of new experimental therapeu­
tic approaches such as resectional surgery in pulmonary
emphysema [45, 46],
Chronic pulmonary vascular occlusion. It has been
shown that V'o2peak is closely correlated to the amount
of functional vascular bed in chronic pulmonary vas­
cular occlusion [22, 47]. However, exercise testing in
such patients carries a significant mortality risk and
should not be performed when there is a history of arr­
hythmias or syncope, or clinical signs of right heart fail­
ure. Furthermore, the power output attainable during
exercise in patients with primary pulmonary hyperten­
sion correlates well with haemodynamic variables mea­
sured at rest [48-50]. Consequently, indications of CPET
in such patients should be carefully established on an
individual basis.
Cystic fibrosis. There is evidence to support the use of
CPET as a tool in the prognosis and management of
patients with cystic fibrosis [51].
Preoperative assessment
Major abdominal surgery in elderly patients. While mor­
tality remains high in elderly patients undergoing major
abdominal surgery, chronological age is a poor guide
to physiological status. It has been suggested that reduc­
tion of cardiopulmonary reserve during exercise can
provide a reasonable prediction of mortality following
major surgery [52, 53]. It has been suggested that there
is a need for a system for grading operative risk that
includes evidence of abnormal cardiopulmonary func­
tion based on CPET.
Preoperative assessment of lung cancer resectional
surgery. Information on predicted postoperative lung
function: 1) helps to modulate the amount of lung par­
enchyma to be resected; and 2) determines the type of
2668
J.
ROCA, B.J. WHIPP
preoperative strategy needed to prevent postsurgical
complications. Resting pulmonary function tests are
considered adequate to evaluate patients with low risk
(forced expiratory volume in one second (FEVl) >2 L
and transfer factor of the lung for carbon monoxide
(TL.CO) within the reference limits) of postsurgical com­
plications [54-59]. However, CPET plays a pivotal role
in the evaluation of patients with moderate to high risk
[57, 58, 60, 61].
Preoperative assessment of lung-reduction surgery,
Lung-reduction surgery [45,46, 62] is the newest thera­
peutic option for patients with emphysema, but still re­
quires further evaluation. As indicated above, CPET plays
a key role in this type of surgery, both in the selection
of potential candidates and to establish the programme
of physical rehabilitation before and after surgical in­
tervention.
Rehabilitation programmes : patient assessment and exer­
cise prescription
Exercise therapy should be part of rehabilitation pro­
grammes that aim to improve both quality of life and
physiological status in patients with COPD and other
forms of chronic lung disease [3, 63-72]. CPET should
always play a central role in the assessment of candi­
dates before the rehabilitation programme and in the
subsequent modulation of the exercise prescription,
whereas simpler tests (i.e. 6 min walk) are useful for
monitoring during the rehabilitation programme.
While it is accepted that physical training can pro­
duce physiological improvements in patients with mod­
erate to severe COPD - such as: 1) increase in exercise
tolerance (V'02peak) [71]; 2) decrease in ventilatory re­
quirements [73]; and 3) faster oxygen kinetics [74] controversy still remains regarding the type of training
programme that is most appropriate [63-65].
Assessment of impairment-disability
It is now well accepted that CPET provides different
and relevant information in impairment-disability eval­
uation [5-7], compared to resting cardiopulmonary mea­
surements [75]. Consequently, CPET constitutes a key
tool in this area. Clarification of this field should be
greatly facilitated with the adoption of the conceptual
framework proposed by the World Health Organization
(WHO) [76]. According to the W HO, the illness-rela­
ted phenomena should be classified in three different
categories: 1) impairment, used to describe loss of func­
tion; 2) disability, corresponds to the resulting reduc­
tion in exercise capacity; and 3) handicap, the total
effect of the illness on the subject's social life.
Lung, heart‘lung transplantation
Lung and heart-lung transplantation are now accep­
ted as viable therapeutic options for patients with end-stage
vascular and parenchymal lung diseases. CPET is a use­
ful tool in the preoperative phase of the evaluation of
the patient's impairment. CPET may also contribute to
the evaluation of disease progression [77], the appro­
priate timing for surgical intervention [78, 79], and can
be used to guide preoperative rehabilitation [48, 8083]. As yet, however, there are no recommendations as
to how indices of exercise performance should be used
in the decision-making process of when a patient should
be selected for a lung transplant. In contrast, such re­
commendations do exist, and have been successfully
applied, in candidates for cardiac transplantation; sur­
vival has been shown to be well con-elated with V'02
peak [78,79].
Several studies have shown that pulmonary and car­
diac function are typically satisfactory in patients with
adequate allograft tolerance. The majority of these
patients report considerable improvement in functional
outcome and life satisfaction, often with resumption of
a normal life style. The available post-transplant data
demonstrate that virtually all recipients have persistent
exercise impairment, regardless of the underlying dis­
ease or type of transplant procedure [83]. This is prima­
rily attributed to skeletal muscle dysfunction [83-85].
In the postoperative phase, CPET is useful in the assess­
ment of limitation of exercise tolerance, discriminating
among its potential causal factors and as a guide to post­
operative rehabilitation.
Protocols
The goal of CPET protocols is to stress the organ
systems involved in the exercise response in a controlled
manner. For this reason the testing generally involves
exercising large muscle groups, usually the lower extre­
mity muscles. A key requirement is that exercise stim­
ulus be quantifiable in terms of the external work and
power performed. Simpler tests, such as step tests or
timed distance walks (i.e., 6 or 12 min walk) can pro­
vide measures of exercise tolerance, but are not as use­
ful in diagnosis as incremental tests [1, 2, 87]. The
purpose of this section is to describe the characteristics
of recommended clinical exercise protocols, both in­
cremental and constant work rate tests.
Standard clinical exercise protocol: incremental test
The appropriateness of the integrated systemic res­
ponses to the tolerable range of work rates are best stud­
ied utilizing incremental exercise testing [88] (fig. 2).
This provides a smooth gradational stress to the sub­
jects so that the entire range of exercise intensities can
be spanned in a short period of time. Technological ad­
vances (see section on equipment) have made it possi­
ble for sufficient density of data to be acquired in a test
lasting less than 20 min (fig. 3), including: 1) mea­
surements at rest; 2) 3 min of unloaded exercise; 3) in­
cremental exercise (approximately 10 min); and 4) 2
min recovery, at least. The recommended increment­
al exercise testing protocol, usually cycle or treadmill
ergometry, is described in detail in the section on proce­
dures . Electronically braked cycle ergometry with con­
stant pedalling frequency, of 60 revolutions per minute
(rpm) for example, is recommended (as discussed in the
section on equipment). Equivalent results are obtained
when work rate is either increased continuously (ramp
CLINICAL EXERCISE TESTING
200
150
20 W'inin'1
CD
e!
100
o
mmm
J.-I
.
-
-
K
.
, ' f
50
10 W-min"1
0-3 -2 -1 0 1 2 3 4 5 6 7 8 9 1 0 y Recovery
Time min
End of exercise
Fig 2. - Graphical representation of standard incremental exercise
protocols. Equivalent results are obtained when work rate is either
increased continuously (ramp test) or by a uniform amount each minute
(1 min incremental test) until the patient is limited by symptoms
(he/she can not cycle >40 rpm) or is not able to continue safely. The
increment rates of 10 W-min-1 (-- ) to 20 W-mhr1 (---) is set
according to the characteristics of the patient in order to obtain approx­
imately 10 min duration of the incremental part of the protocol.
I
incremental exercise
[i.e. approximately 10 min duration;
with an Incrementation rate of 10 to 20 W'tnin"1)
I
Recovery
(Measurements during
2 min of unloaded cycling)
2669
If the ergometer is a motor-driven treadmill, then
Balke's protocol [90, 91] is considered the most appro­
priate for its simplicity. The speed of the treadmill is
kept constant (5-6 km*lrl) during the protocol while
the slope is progressively increased ( 1-2 %-min-1).
Standard nonin vasive CPET carried out breathing
room air (Fl,O2=0.21) involves acquisition of expired
oxygen and carbon dioxide concentrations (F e ,02 and
F e ,C02, respectively), W R, respired airflow, jfc and sys­
temic arterial pressure as primary variables. ECG and
pulse oximetry should be continuously monitored dur­
ing the test. It is useful to establish a sense of the pati­
ent's exercise-related perceptions during the exercise test
and at the point when the subject discontinues exercise.
This includes exertion, dyspnoea, chest pain and skele­
tal muscle effort. Quantifying these perceptions should
be done using standardized rating procedures (Borg
scale, visual analogue scale (VAS), etc.) [92].
Interpretation of CPET results in patients with lung
disease, however, often requires evaluation of pulmon­
ary gas exchange [91]. In these cases, arterial cannulation (preferably radial or brachial) is needed to obtain
partial pressures of respiratory gases in arterial blood
(Pato 2 and Pa,C0 2) and to determine DA-a,0 2 [91, 93].
This also provides information on acid-base status (pH,
Pa,C0 2 and base excess) and allows continuous moni­
toring of systemic arterial blood pressure during the
test. However, while "arterialized venous blood11 (e.g.
from the dorsum of the heated hand) gives good val­
ues for PC02 and pH it is not appropriate for Po2. Fur­
thermore, estimation of arterial respiratory blood gases
through expired oxygen and carbon dioxide profiles or
"transcutaneous" electrodes and pulse oximetry should
not be used as indices of Pa,0 2 and Pa,C02 during exer­
cise [94-96], It is important to recognize that arterial
blood sampling immediately after exercise does not pro­
vide an adequate assessment of blood gas values at peak
exercise. However, while pulse oximetry does not indi­
cate arterial Pa,0 2, it does provide valuable information
on oxyhaemoglobin saturation during exercise.
Other tests
Constant work rate exercise testing. Constant work
rate exercise can result in steady-state responses when
work rate is of moderate intensity (fig. 4). In contrast,
Fig. 3. - Recommended incremental exercise protocol using an elec­
tronically braked cycle ergometer.
test) or by a uniform amount each minute (1 min incre­
mental test) until the patient is limited by symptoms
(he/she cannot cycle >40 rpm) or is not able to conti­
nue safely. The increment size should be set according
to the characteristics of the patient in order to obtain
approximately 10 min duration of the incremental part
of the protocol. This may represent incrementation rates
of 10-20 W-min-1in a healthy sedentary subject or less
in a patient [1, 2, 87]. Modifications in the design of
the protocol and/or the measurements to be carried out
during the test can be considered depending on: 1) the
characteristics of the subject (physically fit or severely
limited); and 2) the clinical problem prompting the in­
dication of CPET (e.g., exercise-induced asthma, as
described in detail by S terk et al, [89]).
Fig. 4. - Time course of oxygen uptake during constant work rate
exercise in one healthy subject (-- ) and one COPD patient (— —).
2670
J. ROCA, B.J. WI-IIPP
a constant work rate of high intensity for the individ­
ual typically results in continually changing values in
most variables of interest. Consequently, attainment of,
or failure to attain, a steady-state V'0 2 during a constant-load test can be used to determine if a particular
task is sustainable by the individual.
However, even for moderate exercise, steady-state
cardiorespiratory responses do not occur instantaneou­
sly when exercise begins; rather, they change over a
period of several minutes, but at rates which are dif­
ferent for each variable (V o2, V'CO2,jfc and VE). It has
been demonstrated that both cardiopulmonary disease
[13] and level of fitness [97] modify the time course of
the nonsteady-state responses to exercise. Hence, the per­
iod of dynamic adjustment to a constant work rate test
provides information regarding the dynamic behaviour
of lung function, haemodynamics and tissue oxygen uti­
lization. However, there is, to date, virtually no infor­
mation on the confidence limits, reproducibility and
predictive value of the derived parameters in patient
populations. Consequently, the utility of quantifying dy­
namic responses to constant work rate exercise in clin­
ical exercise testing remains to be established.
Some factors, however, such as the work efficiency
or the substrate mixture undergoing oxygenation res­
piratory quotient (RQ), require steady-state determina­
tion. It is necessary, in this case, to ensure that variables
are measured only after the dynamic periods of adjust­
ments are completed.
The fc during constant work rate testing is commonly
used as a guideline in exercise prescription (rehabilita­
tion programmes) based on the potential for self-moni­
toring and self-adjusting the exercise dose during training.
Identification of a training effect following an exercise
programme may also be based on the finding of a reduc­
tion in fc for a given constant work rate.
Determining the need for oxygen supplementation dur­
ing exercise or titrating oxygen prescription is most rea­
sonably performed during submaximal constant work
rate exercise, simulating levels of activity needed to
perform daily activities [91]. Furthermore, because pat­
ients with COPD may be limited during exercise by
impaired lung mechanics [2, 19, 26, 27], assessment of
steady-state ventilation, and its pattern, might be of
interest in evaluating the capacity for sustaining a given
task.
It should be recognized that exercise recovery also
follows a dynamic time course. Re-establishment of
resting steady-state conditions following moderate exer­
cise requires a similar time as for reaching the exercise
steady-state, whereas recovery is delayed for consider­
ably longer periods following high-intensity exercise.
Even in young healthy subjects, resting levels of arte­
rial blood lactate concentration, for example, are not
achieved until approximately 1 h after maximum exer­
cise.
Assessment of exercise-induced broncho constriction.
Patients with reactive airways disease may experience
bronchoconstriction following (or during) exercise.
Exercise-induced bronchoconstriction (EIB) is observed
in 70-80% of the population with clinically recogniz­
ed asthma [89, 98, 99]. However, up to 40% of chil­
dren with documented EIB have no symptoms of asthma
[100]. An exercise challenge is, therefore, often worth­
while for patients complaining predominantly of symp­
toms associated with exercise.
Respiratory water loss was identified as the stimulus
by which exercise induces bronchoconstriction in the
late 1970s and it was proposed that evaporative water
loss induced cooling of the conducting airways causing
them to narrow [101]. This later led to the proposal that
EIB occurred as a result of vascular engorgement dur­
ing rapid rewarming of the airways at the end of exer­
cise [102]. Other studies, however, have demonstrated
that the dehydrating effects of water loss are more impor­
tant than the cooling effects. This led to the suggestion
that an increase in airway osmolarity and release of
mediators was the mechanism whereby the airways nar­
rowed in response to exercise [103, 104]. At present,
there is no direct experimental evidence to support either
of these hypotheses.
However, these proposals led to the conclusion that
tests for EIB should aim to achieve a high level of ven­
tilation (V'E 15-22 times the predicted FEVi) while
breathing air containing less than 10 mg-L-1 of water
(20~25°C and <50% relative humidity). Breathing com­
pressed air via a valve will ensure low humidity of the
inspired air. Recording ventilatory and gas exchange
responses during the exercise period allows the level of
Ve and metabolic stress to be documented.
The required V'E can be achieved either by running
on a treadmill or by cycling [105]. Exercise performed
with a cycle ergometer, however, is generally preferred
because: 1) the breathing valve is more easily suppor­
ted; 2) measurement of ventilation is easier; and 3) the
work rate is largely independent of body weight. The
target work rate can be estimated by equations relating
Ve to V o 2 and Vo 2 to work rate. For treadmill exer­
cise, body weight becomes a more important factor for
consideration [106, 107] (table 2). The protocol rec­
ommended consists of a rapid increase in the work rate
to the target value within 3-4 min and sustaining the
target V’E for at least 4 min. For the first, second and
third minutes of cycling the work rate may be conve­
niently set to 60, 75 and 90% of the target value. At
the beginning of the fourth minute the target work rate
is set and Ve carefully monitored [89, 98],
In the patient with EIB, values for FEVi generally
fall to a minimum within the first 10 min after cessa­
tion of exercise, with substantial recovery by 30 min
postexercise, A useful protocol to assess the response
to exercise is to perform forced spirometry before (fol­
lowing recommended standardization) and then FEVi
manoeuvres at 1,3, 5, 7, 10, 15 and 20 min after exer­
cise. The response is assessed as the percentage fall in
FEVi postexercise. The lowest FEVi recorded after
exercise is subtracted from the pre-exercise value and
the difference is expressed as a percentage of the pre-exercise FEVi. A greater than 10% fall in FEVi after exer­
cise is regarded as abnormal and >15% as diagnostic
of EIB. For a drug to be."protective", the fall in FEVi
after exercise should be less than 50% of that docu­
mented on placebo. The coefficient of variation of the
exercise response (percentage fall in FEVi after exer­
cise) is 20% or less when two tests are performed under
the same conditions within one month.
If the test is negative, it is important to check that:
CLINICAL EXERCISE TESTING
2671
Table 2. - Conditions for evaluating the presence and severity of exercise-induced bronchoconstriction (EIB)
Safety
Risk factors or contra-indications for exercise should be assessed when considering patients with EIB.
Oxygen saturation at rest should be >94% and pulse oximetry should be continuously monitored during the test, in addition to
adherence to general guidelines for safety
Measurement
FEVi measurement is mandatory, pre-exercise and 1,3, 5, 7, 10, 15 and 20 min postexercise. In elite athletes, FEF25-75 should
also be measured to assess EIB
FEVi at rest
>75% predicted or 80% of patient's usual value (if known) and reproducible i.e. <10% variation
Medications to be withheld
6 h for short acting bronchodilators, sodium cromoglycate and nedocromil sodium
24 h for long acting bronchodilators, theophylline and antihistamines
No steroids or caffeine on the study day
Factors to be controlled for
EIB >3 h
Infection >6 weeks
Exposure to pollen (test best performed out of season for pollen sensitive subjects)
Type of exercise
Cycling or running. Cycling is recommended as safest and easiest
Intensity of exercise
V'02 sufficient to raise exercise V'E (L-min-1) to 18.5±3.5 times predicted FEVi for at least 4 min
V'E (L-mhr1) = 28 V’02 (L-miir1) + 0.27
V'o2 (mL-min-i) = ¡0 W + 500
For treadmill, V'o2 target will need to be greater than 35 mL-kg^-min-1 to achieve required V'E (see references [98, 106, 107]).
Duration of exercise
6-8 min to allow target V'E to be achieved and sustained for at least 4 min
Inspired air
Compressed dry air is recommended
Room air containing <10 mg J^O -I/1, temperature <25°C, <50% relative humidity is adequate,
Cold air only needed if it represents environmental conditions
Index of severity
If patient not taking inhaled or oral steroids: AFEVl 10-25%: mild
AFEVl 25-50%: moderate
AFEVl >50%: severe
Recovery
Spontaneous or with administration of a (32-agonist.
Supplemental 0 2 should be provided if needed at any time during the test
FEVi: forced expiratory volume in one second; FEF25-75: forced midexpiratory flow; V'o2: oxygen consumption; V’E: minute
ventilation; A: change in.
1) the pre-exercise lung function was not unusually low;
2) the target V'E was achieved and sustained for at least
4 min; and 3) the water content of the inspired air was
<10 mg H 20*L'1. While cooling the inspired air can
increase the severity of the airway response in some
subjects, this same effect appears to be achieved by pro­
longing the duration or increasing the intensity of exer­
cise under temperate inspired air conditions [108]. It
should be noted, however, that exercise itself is not
required to achieve the target ventilation and voluntary
eucapnic hyperventilation of air containing approxi­
mately 5% carbon dioxide can be performed [109,110].
The protocol recommended is appropriate for both
general laboratory assessment of EIB and evaluation
of the protective effect of pharmacological agents [111,
112]. To maximize the likelihood of a positive response,
bronchodilator agents, sodium cromoglycate and nedo­
cromil sodium should be withheld prior to exercise for
a period commensurate with their duration of action
[112, 113]. Furthermore, the subjects should avoid vig­
orous exercise for at least 2 h prior to testing. During
the test, in addition to routine safety precautions requir­
ed for all exercise testing, it is important to have rapid­
ly acting inhaled bronchodilators (e.g. salbutamol)
available to reverse a severe episode of bronchospasm.
Equipment and quality-control programme
Choice of ergometers
Two modes of exercise are commonly employed in
cardiopulmonary exercise tests: treadmill and cycle ergo­
meters. The motor driven treadmill is used to impose a
progressively increasing exercise stress through a com­
bination of speed and grade (slope) increases. Several
incremental protocols are popular [90,114-118]. During
treadmill exercise testing a somewhat larger muscle mass
can be brought to bear than in cycle ergometry, leading
to a greater stress of the organ systems mediating the
exercise response. On average, maximal oxygen uptake
is reported to be 5-10% higher on a treadmill than on
a cycle ergometer [119-122], This may be important to
detect abnormalities (i.e. cardiac ischaemia) that occur
only with the highest metabolic demand. The main dis­
advantage of using a treadmill is the difficulty of quant­
ifying the work rate accurately. The relationship between
speed and grade of the treadmill and the metabolic cost
of performing work is difficult to predict: the "walking
skill" and pacing strategy are important determinants.
Holding on to the treadmill handrails can substantially
2672
J. ROCA, B.J. WHIPP
modify the metabolic cost of treadmill walking, but by
an unpredictable amount. These uncertainties make the
profile of V o 2 increase difficult to interpret. The rela­
tionship of other variables to V o 2, however, remains ap­
propriate.
For laboratory exercise testing, there are several ad­
vantages in using a cycle ergometer. It is generally
cheaper and requires less space. It is less prone to intro­
duce movement or noise artefacts into measurements
(i.e. blood pressure auscultation is generally easier). The
most important advantage, however, is that the rate at
which external work is performed is much more easi­
ly quantified. Although the metabolic cost of moving
the legs at ostensibly zero load is a confounding factor
[2, 123-125], this represents a constant offset as long
as the pedalling cadence is kept constant. The pre­
dictability of the relationship between the imposed cycle
ergometer work rate and the metabolic energy expen­
diture is important for disease evaluation.
The electronically-braked cycle ergometer [126] is
generally used in CPET. Friction-braked cycle ergometers [127] generally provide a work rate that varies
with the cycling frequency. In contrast, the electroni­
cally braked cycle ergometer allows direct quantifica­
tion of the work rate performed, which is usually
independent of cycling frequency. It can also be readily
computer controlled; this allows the work rate to be in­
cremented continuously (ramp exercise) [88,128-132].
Recently, cycle ergometers have become available that
allow "true unloaded'1pedalling (i.e. internal resistance
of the ergometer is overcome by means of a motor
"assist"). A lower starting work rate can therefore be
selected; this is important for the most debilitated patients.
Gas exchange measurements
A central focus of comprehensive CPET is the inter­
pretation of gas exchange responses. Consequently the
methods used to measure V'o2 and carbon dioxide out­
put V'C02 are of important. However, these measurements
are not trivially easy to make, and a clear understand­
ing of the methods involved and the required quality
control procedures are necessary prerequisites.
The technologically simplest technique of measuring
gas exchange involves directing the expired air into a
collection bag by use of a suitable breathing valve. A
timed collection is made; the mixed concentrations of
carbon dioxide and oxygen in the bag and the volume
of the bag are subsequently measured. This allows cal­
culation of V o 2 and V co 2 [91, 133, 134]. By conven­
tion, V102 and V'C02 are expressed under standard
temperature and pressure, dry (STPD) conditions while
ventilation is expressed at body temperature and pres­
sure, saturated with water vapour (BTPS) (for complete
definitions of STPD and BTPS, see Appendix). Use of
highly accurate gas analysers is, naturally, essential.
Though careful attention to technique is crucial, this
method is capable of very precise measurements even
at high metabolic rates. However, it is less suitable for
rapid incremental exercise protocols as it is cumber­
some and the resulting data density is relatively low.
Bag collection systems are now mainly used as a val­
idation technique for more complex gas exchange sys­
tems.
Systems featuring mixing chambers allow continu­
ous measurement of V o 2 and V'C02 [2, 135, 136]. The
subject respires through a breathing valve and expired
air is directed through a baffled chamber. The concen­
trations of carbon dioxide and oxygen are measured
continuously at the distal end of the mixing chamber.
Expired volume is measured, usually breath-by-breath.
V o 2 and V co 2 can then be calculated from the gas con­
centration signals. In the steady-state of exercise, mix­
ing chamber systems are capable of accurate metabolic
measurements. However, since the washout of the mix­
ing chamber requires a finite time (which depends on
the airflow), the volume and gas concentration signals
will be misaligned in the unsteady-state, leading to inac­
curate calculations unless appropriate corrections are
made. For incremental protocols commonly used in clin­
ical CPET, ventilation and mixed expired gas concen­
trations do not change rapidly and the accuracy of a
well designed mixing chamber system may be quite
acceptable (though care must be taken to recognize that
slurring of the response profile is a necessary conse­
quence not only when variables change rapidly but also
when they change direction (i.e., expired partial pres­
sure of oxygen (Pe,02) at the lactate threshold).
With the ready availability of on-line digital com­
puter analysis of physiological transducer signals, it has
become practical to compute V o 2 and V co 2 breath-bybreath [137-141]. Utilizing algorithms first reported in
1973 [137], a signal proportional to expired airflow and
signals proportional to fractional concentrations of car­
bon dioxide and oxygen measured near the mouth are
sampled at high frequency (typically 50-100 Hz). As
shown schematically in figure 5, the breath is broken
down into a number of parts and the V'C02 and V'o2
computed for each interval. The measurements for each
interval are summed over the expiration to compute the
total volume of carbon dioxide exhaled and oxygen
taken up in the breath.
V'C02 = 2 FE£ o 2-VE-At
V'02 = 2 (Fl,02 - FE,02)*VE-At
where VE is the instantaneous expired airflow, At is the
sampling interval and Fl>02, Fe ,02 and Fe,C02 are the
fractional concentrations of inspired oxygen, expired
Fig 5. - Diagram illustrating calculation of breath-by-breath carbon
dioxide output (VC02), which is derived from VC02 = Z^E,C0 2V'E-At V'Exp: expired airflow; FE,C02; expired C02fraction; At; sam­
pling interval.
2673
CLINICAL EXERCISE TESTING
oxygen and expired carbon dioxide, respectively. These
calculations must accommodate water vapour, barometric pressure and ambient temperature variations in
order to obtain STPD values in addition to the "nitro­
gen correction factor" when only expired volumes are
measured (see Appendix). As importantly, compensa­
tion is necessary for the delay between the time at which
gas is sampled at the mouth and the time at which the
gas concentration is measured within the gas analysers
(usually in the order of 0.25 s). Breath-by-breath ana­
lysis, therefore, requires precise knowledge of (and
computerized compensation for) the transport delays of
both gas analysers [140, 142-144]. Thus, the increased
temporal resolution of breath-by-breath analysis comes
at a cost of more exacting quality control requirements.
Measurement of oxygen uptake when Fl,o2 is high
presents particular problems. First, the source of inspir­
ed gas must have a constant oxygen fraction (i.e., from
large gas bags). Further, when F l 02approaches 1, nitro­
gen balance corrections become more and more subject
to error [142]. Depending on the methodology employ­
ed and the accuracy required, Fi,o2of 0.8 or more should
be regarded as beyond the practical upper limits of accu­
rate measurement of Vo2.
concentration functions) is beyond the scope of the pre­
sent document.
One can forego the necessity for serial blood samp­
ling and even, in many cases, enhance the discriminability of LT by utilizing a particular cluster of ventilatory
and pulmonary gas exchange responses, which provide
noninvasive estimation of LT (V-slope method or time
course of ventilatory equivalents for oxygen and car­
bon dioxide; see Appendix). However, the indiscri­
minate use of noninvasive estimators of LT should be
discouraged. Without the supporting evidence that the
more rapid rate of change of Vco 2 or V'E is a conse­
quence of the developing metabolic acidemia, the change
could also reflect any of a range of hyperventilatory
causes of particular concern in patients with lung dis­
ease.
Furthermore, the validly of the noninvasive estima­
tion of LT under "complicating" conditions such as
chronic hyperventilation, progressive exercise-induced
hypoxaemia, and impaired peripheral chemosensitivity
with an associated high airway resistance remains to
be established.
Exercise testing measurement systems
The lactate threshold
Although there is no generally agreed upon proce­
dure for normalizing work intensity, most would con­
cede that the range of work rates within which there is
not a sustained metabolic (chiefly lactic) acidemia (i.e.,
moderate exercise) may be sustained for long periods,
but heavier exercise may not [2]. The threshold V'o2
for arterial lactate concentration increase (the lactate
threshold (LT)) may, therefore, be considered to parti­
tion exercise into moderate and heavy intensities, with
important implications for the ability to sustain a par­
ticular work rate. However, it should be emphasized
that physiological mechanisms underlying the increases
in muscle and blood lactate concentration that occur
above LT remain a highly contentious issue; some in­
vestigators even question whether the arterial lactate
concentration profile actually provides evidence of thre­
shold behaviour [97, 145]. The lactate threshold may,
therefore, be considered to: 1) partition moderate from
heavy-intensity exercise; 2) trigger a series of physio­
logical responses that stress ventilation, pulmonary gas
exchange, and acid-base regulation; and 3) have im­
portant implications for the ability to sustain muscular
exercise, both in normal individuals and in patients with
impaired systemic function.
Determination of LT. The lactate threshold is highly
task-specific, It occurs at an appreciably lower V'02 for
arm exercise than for leg exercise, and is typically lower
for cycle ergometry than treadmill exercise, reflecting
the magnitude of muscle mass over which the work is
distributed, A wide range of techniques have been ad­
vocated for estimation of LT, including both direct mea­
surements and indirect estimation [91]. The description
of the direct measurements of LT [146, 147] (arterial
lactate concentration-pyruvate concentration ratio; log­
arithmic arterial lactate concentration and bicarbonate
High intensity data analysis and display of a range
of relevant variables (e.g. [2]) is most likely to provide
patterns of response that optimize discrimination (espe­
cially of subtle differences). Tests using less complex,
and certainly less expansive equipment and procedures
can also provide useful information, however, especially
if the issue in question is well focused (see [116] for
discussion). Figure 6 presents a schematic representa­
tion of the analytical devices that constitute a recom­
mended measurement system for CPET. The following
components are usually present: 1) digital computer; 2)
flow or volume measurement system; 3) gas analysers;
4) electrocardiograph (ECG); and 5) pulse oximetry.
Measurement of systemic arterial blood pressure and
analysis of respiratory blood gases [93,94] and acid-base
status [2, 148] should also be considered part of the
basic CPET equipment.
Digital computer. The analogue-to-digital converter
needs to sample a number of signals at 50-100 Hz and
the computer should be capable of computations of vari­
ables within each sampling interval. A key function is
*
00
0
Graphical
display
Derived
variables
§
O 3
3 T3
< C
CD
CD
__ ^ Ergometer
control
— ►Data storage
Fig. 6. - Schematic representation of the equipment used for car­
diopulmonary exercise testing (CPET). A/D: analogue/digital.
J. ROCA, B.J. WHIPP
2674
breath detection, often signalled by the detection of the
onset of expiratory airflow. The breath-by-breath val­
ues of a number of variables should be available for
on-line graphical or tabular display and should be stored
for later analysis.
Flow or volume measurement of respired air. A num­
ber of transducers have been used for measurement of
respired flow or volume during exercise. In part, the
expansion of the choice of transducers has resulted
from the use of computerized data analysis. A nonlin­
ear relationship between the actual flow or volume and
transducer output can be accommodated by digital com­
puter processing, as long as the nonlinearity has been
well characterized and is static (i.e., the relationship
does not vary with time). Furthermore, the choice of a
flow versus a volume transducer is no longer crucial
since numerical integration or differentiation can be
employed to calculate one quantity from the knowledge
of the other. A key consideration is whether the trans­
ducer can be positioned near the mouth. Such trans­
ducers are capable of sensing bi-directional flow or
volume. They also eliminate the need for a nonrebrea­
thing valve which means that the apparatus dead space
can be lower. Both the ERS and the American Thora­
cic Society (ATS) have established standards for flow
and volume measurement in the context of spirometry
[149, 150] (table 3). The transducers used in CPET
should also meet these standards. Transducers common­
ly employed for measuring flow or volume in CPET
are listed below.
Pneumotachograph. This flow transducer measures the
pressure drop across a low resistance screen [2 , 119,
151, 152]. Since laminar flow and constant tempera­
ture are required and sputum impaction on the transdu­
cer screen can degrade performance, pneumotachographs
have generally been positioned well downstream from
the mouth. The latter two concerns, however, can be
avoided by measuring inspired rather than expired air­
flow.
Pitot tube flowmeter. This device measures the differ­
ence between pressure at orifices facing the flow stream
and orifices perpendicular to the flow stream. Turbu­
lent (rather than laminar) airflow is involved and the
pressure difference is proportional to the square of the
flow rate [2, 153].
Mass flow meter. This device is related to the older hot
wire anemometer in which a wire is placed in the airstream. The current required to heat the wire to a cer­
tain temperature increases as airflow increases [154,
155]. In one configuration presently used, two wires
heated to different temperatures are utilized; flow detecTable 3. -
tion depends on the fact that the hotter wire loses heat
more rapidly than the cooler wire. Compensation is made
electronically for changes in gas temperature. The sig­
nal generated is (nonlinearly) proportional to the num­
ber of molecules passing the sensor rather than the
volume of gas the molecules occupy.
Turbine volume transducer. A lightweight impeller is
placed in the flow stream and the number of interrup­
tions of a light beam by the impeller are counted by a
computerized system [156]. Although small, the mass
of the impeller causes a small "start up" and "overrun"
at the start and end of the breath, respectively; The dy­
namic non linearities can lead to errors in the timing of
the breath [157].
Gas analysers. Different strategies have been employ­
ed to measure the gas concentrations necessary for
breath-by-breath analysis. One is to use an analyser cap­
able of measuring all the relevant gases (carbon dio­
xide, oxygen and, for some purposes, nitrogen). Mass
spectrometry has most often been used for this purpose.
The other approach utilizes separate analysers for each
gas species. Key requirements are stability and rapidity
of response, The dynamics of analyser response has two
separate components: transport delay (the time required
for the gas to traverse the distance from the sampling
site to the analyser) and analyser response (the kine­
tics of response to a change in gas composition intro­
duced into the analyser). Transport delay (generally in
the order of 0.2-0.5 s, depending on the length of the
gas sampling tube and the gas sampling rate) can be
readily compensated for. The analyser response, often
taking the form of an exponential or sigmoid response
to a stepwise change in gas composition [138,139] must
be kept as short as possible and included in the total
1ldelay". An additional concern is the sensitivity of the
analyser to the partial pressure of water vapour in the
sampled gas. Since water vapour concentration in the
sampled gas can be difficult to predict (principally be­
cause the gas temperature at the sampling point is dif­
ficult to predict), this can introduce substantial errors
in metabolic rate calculations [134].
The mass spectrometer ionizes gas molecules in a
high vacuum environment and then separates them (by
one of several schemes) on the basis of mass to charge
ratio. This enables the time courses of a number of gas
species to be measured. These analysers are linear, often
highly stable and have rapid response characteristics
(analyser time constants of roughly 25-50 ms). They
are usually configured to "ignore" water vapour, yield­
ing dry gas fractions. However, the high cost of mass
spectrometers has inhibited their use in most commer­
cial CPET systems.
Discrete carbon dioxide and oxygen analysers have
Minimum requirements of the equipment
Equipment
0 2 analyzer
C 0 2 analyzer
Flow meter
Cycle ergometer
Range
Accuracy*
0-100%
1%
0-10%
1%
0-14 L'm in-1
3%
0-600 W
2% or 3 W above 25 W
Reproducibility
Frequency response
Test signal
1%
1%
<130 ms
<130 ms
<40 ms
At least, two point calibration
At least, two point calibration
3
L syringe
Dynamic torquemeter
3%
*: linearity within the indicated percentage of full scale for each apparatus.
CLINICAL EXERCISE TESTING
been designed specifically for the demands of cardio­
pulmonary exercise testing. Carbon dioxide analysers,
based on absorption of infra-red light by carbon diox­
ide [158], are common. The oxygen analysers that are
employed are different, based on two principles. In para­
magnetic analysers, the effect of oxygen molecules on
a magnetic field is utilized. In the electrochemical ("fuel
cell") analyser, high temperature reactions between oxy­
gen and substrate are measured. However, these dis­
crete oxygen and carbon dioxide analysers have potential
disadvantages. While the analyser output is usually not
a linear function of gas concentration, computerized
correction for these nonlinearities can be made. Also,
analyser measurements are influenced by water vapour
concentration in the sampled gas. This problem has been
circumvented by using sampling tubing composed of
the polymer Nafion. This polymer contains sulphydryl
residues that absorb water, allowing water molecules to
be transported down its concentration gradient to the
exterior of the tubing. The gas that reaches the analyser
contains little water vapour.
ECG . Serial 12-lead ECG is optimal in subjects under­
going CPET for detection of disease. It is important to
detect ST-segment and T-wave changes consistent with
ischaemia and to define arrhythmias that occur with
exercise [127, 159]. Three-lead ECG recordings for de­
tection of cardiac frequency are suitable in young sub­
jects in whom CPET is being used principally to define
exercise tolerance. A key requirement is that electrodes
should be properly placed using sweat resistant adhe­
sive patches and the detection electronics be specially
designed for movement artefact rejection. Computerized
systems enabling continuous display contribute to test
safety; averaging of ECG complexes can improve de­
tection of ischaemic changes. Extensive description of
electrode placement in 12~lead ECG and requirements
of ECG recording in CPET can be found elsewhere
[160].
Pulse oximetry. The current generation of pulse oxi­
meters detect the variation in transmission of light of
two different wavelengths that occur with arterial pul­
ses in an extremity (usually the finger or ear lobe). Be­
cause oxygenated and reduced haemoglobin transmit
certain wavelengths of light differently, this informa­
tion can be used to estimate arterial oxygen saturation
(iSa,0 2) [94]. While useful and convenient for continu­
ous monitoring [161, 162], several concerns need to be
kept in mind in the context of exercise testing. Pulse
oximeters have limited accuracy (95% confidence lim­
its of ±4-5% as compared to directly measured Sa,o2)
[163]. Some authors have reported that pulse oximeters
tend to overestimate true 5a,o2values [164-166], a parti­
cular problem when test results are being used to pres­
cribe oxygen therapy. On the other hand, poor perfusion
of the extremity (yielding decreased pulsatility), which
may occur in cardiovascular disease, may yield falsely
low readings [167]. Movement and stray light can yield
artefacts and dark skin colour can interfere with signal
detection [168, 169]. Furthermore, the inherent limita­
tions of pulse oximetry must be appreciated [91]. These
devices cannot detect the effects of increased carboxyhaemoglobin (or methaemoglobin); its calculations appro­
2675
ximate the oxygenated fraction of available haemoglo­
bin. An additional disadvantage of pulse oximetry is
that iSa,0 2 rather than Po 2 is measured. Pa,02 (and the de­
rived variable DA-a,0 2) are more relevant in assessing
the effects of lung disease on pulmonary gas exchange.
Systemic arterial blood pressure. Auscultation of blood
pressure becomes more difficult during exercise because
of the increase in ambient noise. Yet detection of exer­
cise induced hypertension (or, less commonly, hypoten­
sion) is an important goal in many circumstances [127].
Automated blood pressure measurement systems have
been developed specifically for use during exercise.
Many operate with the oscillometric method, in which,
as the cuff is automatically deflated in stages, pressure
oscillations induced within the cuff by pulsations in the
arm are detected [170]. Despite algorithms designed to
decrease the effects of artefacts, blood pressure mea­
surements may be inaccurate when, for example, the arm
moves during the measurement cycle. Periodic checks
against manual determinations are important.
For studies in which a radial or brachial artery cathe­
ter is inserted to facilitate blood sampling, it may be
useful to measure blood pressure directly. It should be
appreciated that there are (modest) systematic differ­
ences between auscultated and intra-arterial blood pres­
sure measurements [121, 171]. Miniature transducers
that can be attached to the arm while the subjects exer­
cises are available. Meticulous attention to technique
(/.£., exclusion of air bubbles) is necessary, as elsewhere
with CPET, to ensure a good frequency response. Ste­
rility concerns have led some laboratories to use single-use disposable blood pressure transducers, Proficiency
guidelines for the analysis of arterial blood gases should
be followed.
Quality control programme
For CPET systems purchased as a unit, the manufac­
turer should be required to bear responsibility for demons­
trating that the system is capable of accurate measurement.
This might include description of bag collection com­
parisons over a range of metabolic rates and with a vari­
ety of breathing patterns. Algorithms used for breath
detection and for calculation of variables should be pro­
vided to the user. However, it must be stressed that the
user bears the responsibility for ensuring that measure­
ments are accurate and reproducible on a day-to-day
basis. CPET, especially when it features breath-by-breath
gas exchange analysis, requires meticulous attention to
calibration procedures to ensure accurate and repro­
ducible measurements (table 3). A good practice is to
maintain a calibration log book so that long term trends
can be monitored and sudden changes recognized and
addressed.
Daily calibration begins with the determination of
ambient barometric pressure, temperature and relative
humidity. Several other calibration manoeuvres are per­
formed daily (or more often if many tests are done) to
check the operation of key transducers. Most commer­
cial exercise systems facilitate these calibrations.
Verification of calibration of the airflow or volume
transducer can be performed with a precision large vol­
ume syringe (e.g. 3 L). A wide range of flow rates
J. ROCA, B.J. WHIPP
2676
should be performed to simulate the range of flows that
occur between rest and heavy exercise; syringe strokes
varying from 1-10 s in duration cover most of this range.
Agreement in calculated volumes to within ±3% signi­
fies adequate performance [149, 150].
Although the output of most carbon dioxide and oxy­
gen gas analysers is a nonlinear function of gas con­
centration, electronic algorithms aim to create linear
outputs over the desired operating range. For carbon
dioxide, this is usually 0-8%; for oxygen it is 13-21%
(unless testing with hyperoxic gas mixtures is inten­
ded). At least daily, a minimum of two point calibra­
tions of each analyser (with at least two precision gas
mixtures) should be performed. On occasion, certified
tanks with other relevant compositions should be used
to verify linearity. This is conveniently done with: 1)
one 3% carbon dioxide and 13% oxygen tank; 2) one
6% carbon dioxide and 17% oxygen tank; and 3) one
0% carbon dioxide and 21% oxygen tank. Few labora­
tories nowadays are capable of performing Scholander
analysis [172]; mass spectrometry is a reasonable "gold
standard" measurement modality. A good practice is to
maintain a single precision gas cylinder for occasional
use; such a tank can last a number of years and (if reg­
ularly rotated to minimize settling effects) provide a
long-term validation of calibration accuracy.
A third routine calibration must be performed in sys­
tems featuring breath-by-breath gas exchange measure­
ments. The transport delay between the gas sampling
point and the gas analysers needs to be known with pre­
cision so that the respiratory airflow and gas concen­
tration signals can be properly time aligned. A solenoid
allows an abrupt switch between two gas sources with
different oxygen and carbon dioxide compositions and
the time delay can be measured between the solenoid
activation and the detection of change in gas analyser
output. To this delay is added one time constant if the
analyser response is exponential or the half time if the
response is sigmoidal.
It is useful to perform an overall check of the venti­
lation and gas exchange measurements on a regular basis.
One approach is to use a gas exchange simulator [173]
that features a reciprocating piston; injection of a pre­
cision gas mixture at a precisely metered rate yields
simulation of a known V'E,
and V'o2. Day-to-day
variation of these calculations and the variation with
changes in pump rate should be roughly in the range
of ±3%. It should be noted that this gas exchange sim­
ulator does not simulate the moist exhalate at the expired
temperature. Thus, assumptions about temperature and
humidity corrections are not tested.
Other calibration procedures also need to be perform­
ed. Blood pressure transducers can be calibrated with
a mercury manometer. Devices are available to calib­
rate electromagnetic ally braked cycle ergometers [174176]; calibration should be performed at least every 6
months or whenever the cycle ergometer is moved (jar­
ring often disturbs the calibration) or when an unusual
response profile raises concerns about the equipment.
For treadmills, belt speed should be verified by timing
revolutions of the belt with a subject on the treadmill;
accuracy of the grade indication should also be vali­
dated [127]. There are two other quality control checks
that are advisable. A healthy member of the laboratory
Vco2
staff (with a consistent diet) should perform a constant
work rate test at regular (perhaps weekly) intervals.
Steady state values for V'02 differing by more than per­
haps 7% from previous values should engender a thor­
ough system-wide reassessment.
Finally, timed expired gas collections made during
the steady state of exercise can be used as a "gold stan­
dard" to validate ventilation and gas exchange mea­
surements. Though laborious, when performed carefully,
this method is generally accurate to within 2-3%.
Personnel and testing procedures
Personnel qualifications
Cardiopulmonary exercise testing should be conduc­
ted only by adequately trained personnel with a basic
knowledge of exercise physiology. Technicians familiar
with normal and abnormal responses during exercise
and trained in cardiopulmonary resuscitation (CPR)
should be present throughout the test. CPET should be
performed under the supervision of a physician who is
appropriately trained to conduct exercise tests and in
advanced CPR. The degree of subject supervision need­
ed during the test can be determined by the clinical sta­
tus of the subject being tested and the type of exercise
protocol. While it is preferable for the physician to be
present during the test, if it is not possible, he/she must
be readily available to respond as needed. Additional
roles for the physician are the evaluation of the patient
immediately before the test and the interpretation of the
results.
Patient preparation
At the time of scheduling, the subject should be in­
structed to adhere to his/her usual medical regimen; he/
she should not eat for at least 2 h before the test, avoid
cigarette smoking and caffeine, and dress appropriate­
ly for the exercise test. A brief history (with detailed
inquiries about the medications) and physical exami­
nation should be taken to rule out contra-indications to
testing. Results of recent resting pulmonary function
tests (as a minimum forced spirometry) should be avail­
able for patients in whom pulmonary disease is sus­
pected.
On arrival at the CPET laboratory, a detailed explana­
tion of the testing procedure and equipment should be
given to the patient, outlining risks and potential com­
plications as described below. The subject should be
told how to perform the exercise test and the testing
procedure should be demonstrated, if needed. The patient
should be encouraged to ask questions to reduce any
anxiety. The patient needs to become familiar with the
equipment [91]. If the treadmill is used, time is pro­
vided for several practice trials of starting and stopping
until the patient feels confident. If the cycle ergometer
is used, the seat height is adjusted so that the subject's
legs are almost completely extended when the pedals
are at the lowest point and the cycling rhythm prac­
ticed .
CLINICAL EXERCISE TESTING
If arterial blood sampling is required, a catheter is in­
serted into a distal artery (preferably the radial or brachial
artery) of the nondominant arm following proper pro­
cedures to ensure collateral circulation. Sterile techni­
que must be utilized during the catheter placement [148,
177]. A continuous flush device must provide a slow
infusion of a heparinized saline solution. The dead space
of the catheter used for sampling should be less than 1
mL. To avoid spurious dilution, 2 mL of blood is dis­
carded before collecting each arterial blood sample.
Before the test, the ECG electrodes are carefully placed
and secured after preparing the skin to ensure good
recordings (if necessary, the area of the electrode place­
ment should be shaved). A sphygmomanometer cuff is
placed on the upper arm. The mouth piece and nose clip
are then tried and the position adjusted until a com­
fortable position is adopted. The patient is informed that
it is acceptable to swallow with the mouth piece in place
and that he/she must signal any unexpected difficulty
by the signal "thumbs down". The patient is advised to
point to the site of discomfort if chest or leg pain is
experienced.
During the test, the patient is encouraged to carry on
with a regular pedalling rhythm. Use of a metronome
to assist in maintaining rhythm is often helpful. Symp­
toms and degree of discomfort are checked periodical­
ly (see safety precautions below). Good communication
with the patient throughout the whole procedure incre­
ases the subject's confidence and predisposes to good
effort.
During recovery, the patient is told to continue to
pedal, without external workload (or walk at a slow pace
on the treadmill), for at least 2 min during recovery in
order to prevent fainting and to accelerate lactate remov­
al. At the point when the subject discontinues exercise,
after removal of the mouthpiece, the physician should
ask for symptoms (type and intensity) that prompted the
patient to stop exercise. If blood gas analysis is perfor­
med, a last blood sample is taken after 2 min of recov­
ery.
If the test does not provide adequate diagnostic in­
formation because of premature termination or inade­
quate co-operation of the patient, it should be repeated
after a resting period of 30-45 min.
After removal of the radial (or brachial) artery cathe­
ter, adequate pressure must be applied for sufficient time
to avoid bleeding. This can take up to 10-15 min. A
compression bandage should be applied to the site of
Table 4. -
2677
the puncture. The patient is advised not to remove the
bandage and not to use his arm for heavy exercise with­
in 6 h after the test.
Safety precautions
Although CPET may be considered to be a safe pro­
cedure, risks and complications have been reported. A
recent review [160] summarizing eight studies of esti­
mates of sudden cardiac deaths during exercise testing
revealed rates from none (four studies) to five per 100,000
tests. A French study [178] based on 458,000 exercise
tests (1975-1985) reported one death per 76,000 exer­
cise tests. However, the relative risk of an adverse event
is strongly related to the underlying disease, particularly
in subjects post-myocardial infarction or with malignant
arrhythmias. Good clinical judgement should be para­
mount in defining indications and contra-indications for
exercise testing [159]. Table 4 lists absolute and rela­
tive contra-indications to CPET. Cardiac (bradyarrhyth­
mias, ventricular tachycardia, myocardial infarction, heart
failure, hypotension and shock) and noncardiac (mus­
culoskeletal trauma, severe fatigue, dizziness, fainting,
body aches) complications of CPET have been repor­
ted. Consequently, during the test, the personnel should
be alert to any abnormal event.
Indications to stop the test must be clearly established
and known by all the personnel involved in testing.
These indications include symptoms such as: 1) acute
chest pain; 2) sudden pallor; 3) loss of co-ordination; 4)
mental confusion; and 5) extreme dyspnoea. The signs
include: 1) depression of ST segment greater than 0.1
mV (less specific in females); 2) T-wave inversion; 3)
sustained ventricular tachycardia; and 4) fall in systolic
pressure either below the resting value or about 20
mmHg below its highest value during exercise testing.
Relative indications to stop the test are: 1) polymorphic
and/or frequent premature ventricular beats; and 2) hy­
pertension (>250 mmHg systolic; >130 mmHg dias­
tolic). If the exercise test has been stopped for one of
these reasons, the patient should be monitored in the
CPET laboratory until symptoms or ECG modifications
have completely cleared. Admission to hospital for longer
observation or more often for complementary investi­
gation will be necessary in very rare cases. If neces­
sary, intensive care can be administered on site. Full
CPR equipment should be available in the CPET labo­
ratory .
Contra-indications for cardiopulmonary exercise testing
Absolute
Relative
Acute myocardial infarction (3-5 days)
Unstable angina
Uncontrolled arrhythmias causing symptoms or haemodynamic
compromise
Active endocarditis
Acute myocarditis or pericarditis
Symptomatic severe aortic stenosis
Uncontrolled heart failure
Acute pulmonary embolus or pulmonary infarction
Acute noncardiac disorder that may affect exercise performance or
be aggravated by exercise (i.e. infection, renal failure, thyrotoxicosis)
Thrombosis of lower extremities
Left main coronary stenosis or its equivalent
Moderate stenotic valvular heart disease
Electrolyte abnormalities
Severe untreated arterial hypertension (>200 mmHg
systolic, >120 mmHg diastolic)
Significant pulmonary hypertension
Tachyarrhythmias or bradyarrhythmias
Hypertrophic cardiomyopathy
Mental impairment leading to inability to cooperate
High-degree of atrioventricular block
"Relative" contra-indications can be over-ruled if the benefits outweigh the risks of exercise.
J. ROCÀ, B.J. WHIPP
2678
Interpretative strategies
As mentioned above, the main general goal of CPET
is to evaluate the degree of limitation of exercise tol­
erance and to identify its causal factors. Optimal clin­
ical application of CPET should be attained by properly
accomplishing each of the steps indicated in figure 1:
1) identify the clinical problem for which CPET is
required; 2) choose the exercise protocol; 3) ensure high
quality results; 4) present results using a proper format;
5) select adequate reference values to establish patterns
of abnormal response; and 6) compare with character­
istic patterns of certain diseases.
In this section on interpretation of clinically orient­
ed CPET, we first recommend a logical strategy to
approach the problem. Comparative analysis of differ­
ent sets of available reference values is also provided
and, finally, detailed information on the variables indi­
cated in table 5 (definition, information content, units
and calculations) is described in the Appendix.
General guidelines
Several points should be considered in the interpre­
tation of CPET results. First, suggestions in the litera­
ture supporting interpretative strategies based on single
key measurements to direct the flow-chart decision did
not prove to be adequate. The greatest diagnostic poten­
tial and impact on the clinical decision making process
Table 5. - Important variables and recommended plots
Category
Variables
Mechanical work
Gas exchange
Ventilation
Respiratory blood gases
Cardiovascular
Acid-base status
Work rate (W)
Vo2, Vco2, R ER , LT
V 'E , VT,/, VR
Pa,02, Pa,C02, DA-a,02, VD/VT, 5a,02
/C, H RR, Pa,sys, ECG, 0 2 pulse
pHa, Pa,co2, base excess or
standard C 0 3H~ concentration
Dyspnoea, leg pain, chest pain
Symptoms
Basic plots
V'C02 (ordinate) vs V o 2 (abscissa)
Ve/V'co2 (ordinate) and V'e/V'o2 (ordinate) vs V o 2 (abscissa)
Pet,co2 (ordinate) and Pet,o2 (ordinate) vs V o 2 (abscissa)
RER (ordinate) vs V o 2 (abscissa)
V o 2 (ordinate) vs work rate (abscissa)
V e (ordinate) vs
(abscissa)
V'co2 (abscissa) or Ve
(ordinate) vs V o 2
fc (ordinate) and V c o 2//c (ordinate) vs V o 2 (abscissa)
V t (ordinate) vs V’E (abscissa)
Examples of the basic plots are shown in figure 7. V o 2: oxy­
gen consumption; V c o 2: carbon dioxide production; RER:
respiratory exchange ratio; LT: lactate threshold; VR: venti­
latory reserve; ƒ: frequency; Pa,02: arterial oxygen tension;
DA-a,o2: alveolar-arterial oxygen difference; Vd: dead space
volume; Vt: tidal volume; Sn,o2: arterial oxygen saturation;
fc: cardiac frequency; HRR: heart rate reserve; Pa,sys; sys­
temic arterial pressure; ECG: electrocardiogram; pHa: arterial
pH; Pa,co2: arterial carbon dioxide tension; V'E: minute ven­
tilation; P e t ,c o 2: end tidal partial pressure of C 0 2; P e t ,o 2:
end tidal partial pressure of 0 2.
should rely not on the utility of any one individual mea­
surement, although some are obviously more important
than others, but rather on their integrated use [1], Con­
sequently, a key point is the identification of a cluster
of responses characteristic of different diseases.
The major portion of the interpretation strategy is
focused on CPET results generated during maximal,
symptom-limited, incremental exercise testing. This is
currently the most popular, albeit not exclusive, proto­
col. Often, insufficient attention is paid to trending phe­
nomena as the work rate progresses from submaximal
to peak levels. To facilitate this type of analysis, the
results should be formatted in an appropriate manner.
While various display formats can be helpful in this re­
gard (see [2] for example), one suggested format is pre­
sented in figure 7, which displays data obtained in a
normal subject performing cycle ergometry, using an
ergometer that utilizes an "assist" to provide an actual
0 W work rate at "unloaded" pedalling. Figures 7a-d
provide, in addition to the peak V o2, the variables com­
monly used to provide an indirect estimation of the LT
(as described in the Appendix using a cluster of rele­
vant responses for the estimation, rather than relying
on a single index). That is, the V o 2 at which the tran­
sition between moderate and heavy-intensity exercise
occurs is identified. Figure 7e (V o 2 versus work rate)
reflects the exercise efficiency (see section on Response
to exercise in lung disease, subsection on Pulmonary
gas exchange) and the limits of exercise tolerance of
the subject. Figure 7f (ventilation versus V'C02) and fig'
ure 7h (VT versus ventilation) characterize aspects of
the ventilatory response during submaximal and maximal exercise. However, some investigators find the rela­
tionship between Ve and V o 2 during such tests to be
useful. Finally, figure 7g is informative with respect to
the characteristics of the haemodynamic response to
exercise. The next step is to choose adequate reference
values to establish patterns of normal or abnormal res­
ponse. Available data on reference values and present
limitations in this particular issue are discussed below.
This graphical analysis (fig, 7) can be even more illus­
trative of the subject's exercise performance if the cor­
responding reference values are also displayed in each
plot (see [2] for example). A proper interpretation of
CPET results should rely on both numerical analysis
of the variables of interest (table 5) and assessment of
trending phenomena throughout the exercise test using
the graphical analysis suggested above. Comprehensive
CPET can provide extremely useful information in the
judgement on the system's tolerance to exercise by help­
ing to answer the following basic questions in a given
patient: 1) to what extent is the system constrained or
limited? 2) how is it perceived? 3) are the metabolic
requirements for a given task met? and 4) what is the
cost of meeting the requirements?
The final step in this approach to the interpretation
of CPET results is to identify clusters of responses cha­
racteristic of different diseases. It should be pointed out
that each step of the proposed interpretative strategy is
important; unless a flow-chart logic (e.g. [2]) is fol­
lowed, the analysis of CPET results is prone to misin­
terpretations. Different algorithms to guide the decision
making process, in discriminating a range of function­
al abnormalities, have been reported [1, 2 , 87] and are
CLINICAL EXERCISE TESTING
a)
2.5■
■
2.0E
2679
«
1.5*1
CM
o
o
1.0-
■
i
0.50
0
b)
0.5
1.0
1.5
V'o* L-mirr1
2.0
55
5045-
S
O
S
u3
CM
to*
f) 100-
-55
40■
3530«
-50
*
■
■
■
75a
□
°S ■*
P
D
CO
oaa
■3 ■
D
-45
_G
□“ □
□
□
m\
□ft
□
é »■v*\» ■
*
25-
n
-40
O
O
S
LÛ
e
£
./
50-
UJ
S
-35
25-
*.
-30
c) 135-
g) iso
g
CT>
CD
X
i
E
E
CM
o
o
CM
o
ÉÉ
C
O
CD
a
Q_ a
_Q
42
E
45
£f
0.5
1.0
1.5
l/'Og L-min"1
1.0
1.5
V'Oo L-min-1
h) 2.0-
■■
■
1.5 -
£
y
1.0 -
0.5-
0
in*1
V’o2 L-min
m»
0
■■
■
*
25
50
75
V'E Lmirr1
100
Fig. 7. - Basic plots recommended for the interpretation of CPET. Figures a-d, in addition to peak V 0 2, provide the variables commonly used
to indirectly estimate the LT (as described in the appendix). That is, the
at which the transition between moderate and heavy-intensity exer-
legend to table 5.
J. ROCA, B.J. WHIPP
2680
available in the literature. The logical flow commonly
begins with the judgement of whether exercise toler­
ance was normal or not. If the subject shows abnorm­
al limitation to exercise, the next logical step might be
the analysis of the submaximal exercise region in order
to identify the LT. If the physiological changes indi­
cating the transition from moderate to heavy-intensity
exercise occur within the lower end of the expected
range of V'02 for a normal subject (but within it), then
the dilemma is whether the subject had voluntarily
stopped the test early or whether a physiological lim i­
tation to exercise performance can be identified. For
example, were the analysis of the ventilatory function
to show that the ventilatory reserve (VR; see Appen­
dix) was zero or low but the heart rate reserve (HRR)
was high at the end of the test, this suggests that the
patient probably had a ventilatory limitation to the exer­
cise, rather than stopping deliberately or as a result of
acute discomfort, such as that of angina.
The above recommended guidelines for CPET inter­
pretation propose a general logic framework for the ana­
lysis of limitation of exercise tolerance, e,g: cardiovascular
(heart, systemic circulation and blood); respiratory (ven­
tilation, pulmonary circulation and gas exchange); and
peripheral muscle factors (neuromuscular-related ab­
normalities, impairment of muscle microvasculature, ab­
normal cellular oxygen or other substrate utilization).
However, additional factors, such as psychogenic ones,
can also be limiting. However, results of CPET, like
other physiological tests in cardiopulmonary medicine,
demonstrate patterns of abnormality with overlap bet­
ween the responses of patients with different disorders.
Table 6 provides a description of the most characteris­
tic features of abnormal responses in different patient
categories. In each of these categories there is one or
more variables that have a predominant role in deter­
mining the severity of impairment. An expanded and
detailed description of these patterns of abnormality can
be found in elsewhere [85] .
Relatively few studies have evaluated the sensitivity,
specificity and predictive value of patterns of measure­
ments in distinguishing different clinical entities. Even
more importantly, the precise role of clusters of vari­
ables commonly used in the decision making process in
Table 6.
well identified diseases (i.e., evaluation of ILD, preop­
erative evaluation for resectional lung cancer surgery,
etc.) is insufficiently known. For the future, studies ad­
dressing the use of likelihood ratios [179] might be even
more useful to clinicians than sensitivity and specifi­
city, since likelihood ratios refer to actual test results
before disease status is known. This shift to an evidencebased approach [180] for CPET interpretation will, hope­
fully, provide important answers to clinically relevant
questions that are not immediately available,
Reference values
As indicated above, selection of appropriate reference
values is an important step to establish patterns of nor­
mal or abnormal response to exercise stress. An initial
analysis of available data on healthy subjects [1, 2, 12,
118, 181-191] clearly indicated that only five of these
studies [2,181,184-186] fulfill minimum requirements
to be considered as candidates to be used in the clinical
setting [85]. However, Blackie et ah [184] cover a lim­
ited age span (55-80 yrs), and Bruce et al. [186] pro­
vide data obtained with treadmill in a population of
physically fit people. Hence, the analysis of potential
studies in healthy sedentary people providing predic­
tion equations for peak V'02 obtained with incremental
cycling exercise testing is reduced to three sets [12,181,
185], Basic characteristics of these three studies are summ­
arized in table 7. Reference values estimated by Fairbarn
et al. [185] are consistently higher than those provided
by Jones et al, [181]], both in men and women. The
predicted values of Hansen et al. [2, 12] are closer to
either Jones et al, [181] or Fairbarn et al. [185] depend­
ing upon the values of height-weight of the subject in
whom the equations are used. The limitations of the
presently available prediction equations for peak V o2
(and peak work rate) clearly impose limitations on the
interpretative strategy. Moreover, except for fc in the
study of Fairbarn et al. [185], the profile of response
in healthy sedentary subjects (i.e., from submaximal to
peak exercise results) are not available. Furthermore,
adequate prediction equations for even the most impor­
tant variables (i.e., these indicated in table 5) obtained
Patterns of abnormal response to exercise in different diseases
COPD
ILD
PVD
Obesity
Deconditioned
Heart failure
Reduced
Reduced
Reduced
Reduced
Normal for
ideal weight
Reduced
Reduced
LT
Indeterminate,
normal or low
Normal
or low
Low
Low
Normal or low
Low
VEjeserve
Reduced or
none
Normal or
reduced
Normal
Normal
Normal
Normal
HRR
Normal
or increased
N ormal
or increased
Normal
N ormal
Normal
Reduced
or normal
pulse at peak exercise
Reduced
Reduced
Reduced
Normal
Reduced
Reduced
Fall in Sa,02 at peak exercise
Present
or absent
Present
Present
Absent
Absent
Absent
V’lo 2,peak
0 2
(% predicted)
COPD: chronic obstructive pulmonary disease; ILD; interstitial lung disease; PVD: pulmonary vascular obstructive disease; V o2,peak:
oxygen uptake at peak exercise; LT: lactate threshold; Ve,reserve: see Appendix; HRR: heart rate reserve (see Appendix); 0 2
pulse: V o 2,peak/peak cardiac frequency; Sa,o2: oxyhaemoglobin saturation in arterial blood.
CLINICAL EXERCISE TESTING
Table 7. -
2681
Characteristics of the main sets of prediction equations
First author [ref]
Gender
Equations
Population
Subjects
Equipment
H ansen [2, 12]
M
F
V o 2 ,peak (mL-mhr1) (50.75 - 0.372A)Wt
V o 2,peak (mL-mnr1) (22.78 - 0.17A)(Wt+43)
77M
Age 34-74 yrs
Shipyard
workers
Cycloergometer
Breath-by-breath
10-30 W increments
J ones [181]*
M
F
M
F
V'o2,peak (L-min'1) 0.046H - 0.021A - 4.31
V,02)Peak (L-min-1) 0.046H - 0.021A - 4.93
W R (kpnvmiir1) 204H - 8.74A - 1909
W R (kprrvmin-1) 20.4H - 8.74A - 2197
50M/50F
Age 15-71 yrs
V olunteers
Cycloergometer
10 L mixing box
100 kpm-mnr1
increments
B lackie [184]
M
V o 2,Peak (L-min-1) 0.0142H - 0.0494A
47M/8 IF
+ 0.00257Wt + 3.015
Age >55 yrs
V'02,Peak (L-min-1) 0.0142H - 0.0115A
+ 0.00974Wt + 0.651
W R (kprn-mhr1) 6.1H - 26.1A + 0.04Wt + 1.704
W R (kpm-min'1) 74H - 13.0A 4- 3.78Wt + 52
Volunteers
Cycloergometer
10 L mixing box
100 kpnvmnr1
increments
V o 2,peak (L-min-1) 0.023H - 0.031 A
+ 0.0117Wt- 0.332
V'o2,peak (L-min-1) 0.0158H - 0.027A
+ 0.00899Wt + 0.207
Volunteers
Cycloergometer
8.5 L mixing box
100 kpm*min-1
increments
F
M
F
Fairbarn [185]
M
F
111M/120F
Age 20-80 yrs
M: male; F; female; A: age in years; H: height in centimetres; Wt: weight in kilograms; WR: work rate; kprn: kilopondmeters.
For further definitions see legend to table 6. Hansen et al. [2, 12] suggest alternative equations for those subjects with differ­
ences between actual and ideal body weight. *: equations simplified from, Jones et al. [181] provide prediction equations for
cardiac frequency and 0 2 pulse at peak exercise; Fairbarn et al. [185] provide equations for cardiac frequency during submaximal and peak exercise.
Table 8. -
Traditional versus SI units
Traditional
units
Name
Mechanical work
Work rate
Gas exchange
V'Ot
V'C02
RER
Ventilation
V'E
Vr
ƒ
VR
Respiratory blood
Pa,02
Pa,C02
DA-a,02
Vd /Vt
Cardiovascular
/c
HRR
V'oJfc
Acid-base status
Base excess
or bicarbonate
Lactate
concentration
SI
units
Units
Conversion
factor
Units
Name
Sl/traditional
kg-m-miir1
or kpnvmnr1
Watts
W
0.163
mL-min-1
mL-miir1
n'o2
ïî co2
RER
mmol-min-1
mmol’min-1
22.40-1
22.46-1
1.0
L-min-1
mL or L
breaths-miir1
V'E
Vt
ƒ
VR
L-min-1
mL or L
breaths-min-1
1.0
1.0
1.0
1.0
Pa,02
Pa,C02
DA-a,02
VüfVr
kPa
kPa
kPa
at*
0.133
0.133
0.133
1.0
fi
HRR
n'oJfC
beats'mhr1
1.0
1.0
22.4-1
gases
mmHg
mmHg
mmHg
beats-miir1
mL
goal in this area is to stimulate a multi­
centre study, with the aim of generating
an appropriate set of reference values. This
would establish both the magnitude and
profiles of response of the essential CEPT
variables utilizing standardized procedures
and rigorous quality control (within and
among participating laboratories). Addition­
al potential gains from such a programme
would be the comparison among incre­
mental cycling, incremental treadmill and
constant work load exercise testing in a
subset of subjects. Such a project would
provide a major advance in r
the
ability to interpret CPET.
Appendix
SI units
SI stands for "Système International
d'Unités" [192, 193]. These units are the
mmol
result of over a century of international
mM
1.0
mEq-L'1
Base
co-operation to develop a universally ac­
excess
ceptable system of units of measurement.
mM
Lactate
mEq-L-1
1.0
The SI is an outgrowth of the metric sys­
concentration
tem that is being increasingly used all over
For definition of abbreviations see legend to table 5. n'o2, n'C02: millimoles per the world, but which has had rather little
minute of oxygen and carbon dioxide, respectively; kpm: kilopondmeters; Con­ impact in the United States of America
version factor; Traditional unit to SI unit (SI/traditional unit).
(USA), even though Congress passed the
Metric Conversion Act in 1975, which
from the same group of reference subjects are not cur­
endorsed the SI scale. Formally, the SI units were adop­
rently available. The absence of appropriate reference
ted by all the members of the European Union about
values can naturally lead to miscategorization and pos­
20 yrs ago. Table 8 indicates both nomenclature and
conversion factors of the SI units for the variables used
sible misinterpretation of the results.
in the present document.
The Task Force members consider that an important
J. ROCA, B .J. WHIPP
2682
Variables: definitions and calculations
Mechanical work. Work rate (W). This indicates the
amount of work performed per unit of time. Work (kgm 2,s~2, J), in turn, is a physical quantification of the
force (kg*m2'S‘2, N) operating on a mass (kg) that caus­
es its change of position (distance expressed in meters).
The work rate (or power) is measured in watts (kg’m2S‘3 or J’S“1) ’
Gas exchange. Oxygen uptake (V o2, in mL-min-1 or
L-min-1). Is the difference in oxygen flow between in­
spired and expired gas (in STPD conditions, see below).
During steady-state conditions, oxygen uptake and con­
sumption (amount used by the body metabolism in a
given period of time) are equivalent. V'o2 may be cal­
culated as:
Vo2= V'vFl,02 - VF
E’FE,02
or, if only V e is determined:
Vo,=
/ F e ,n 2 v
(F i ,02- ( — — ) - F e ,o 2)-V'e
F ijsu
where V'E is expiratory minute ventilation and V i is
inspiratory minute ventilation, Fi,02 and F e,0 2 are the
fractional concentrations of inspired oxygen and expir­
ed oxygen, and Fi,N2 and Fe,N2 are the fractional con­
centrations of inspired and expired nitrogen.
Carbon dioxide output (Vco 2mL-min"1)* In the absence
of inspired carbon dioxide, V'C02 is the flow of carbon
dioxide exhaled from the body into the atmosphere, ex­
pressed in STPD conditions (see below). During steady
state conditions, carbon dioxide output equals produc­
tion by the body. Calculation of Vco2 can be comput­
ed as the total volume of carbon dioxide exhaled in a
given period of time:
V c o 2 = Fe ,c o2*Ve
where F e ,C02 is the fractional concentration of expir­
ed carbon dioxide.
Respiratory exchange ratio (RER or R, dimensionless).
This is the ratio of V'C02to V o2. RER reflects not only
tissue metabolic activity, but also the influence of trans­
ient changes in body stores of respiratory gases (oxygen
and, most importantly, carbon dioxide). The respirato­
ry quotient (RQ) is the ratio of V'C02 to V o 2 and reflects
the metabolic substrate utilization. During hyperven­
tilation, RER exceeds RQ because additional carbon
dioxide from the body stores is exhaled, whereas the
RER is less than the RQ during transient hypoventila­
tion when carbon dioxide is retained in the body stores.
Lactate threshold (LT, mL'rnhr1), This is the exercise
V'o2 above which a net increase in lactate production
results in a sustained increase in blood lactate concen­
tration. The LT provides: 1) an index of the functional
status of the respiratory-circulatory-metabolic integra­
tion that allows exercise to be sustained aerobically; 2)
an index of sustainability for that particular task; 3) a
frame of reference for optimizing training protocols and
monitoring physical training, rehabilitation and drug in­
terventions; and 4) a component of decision-making stra­
tegies for elucidating the dominant system(s) responsi­
ble for exertional dyspnoea, and exercise intolerance.
Its limitations, however, in patients with lung disease
have been indicated in the text and are further analysed
below:
1) Direct measurements of LT. The response profile of
arterial lactate concentration versus Vo 2 is often not the
most sensitive estimator for LT, as there may be no
clear "break-point" in the profile that can be identified
with sufficient confidence (for some investigators, at
least).
2) Arterial lactate-pyruvate concentration ratio. As a
component of the arterial lactate concentration, incre­
ase reflects pyruvate-dependent increase, as a result of
increased glycolytic flux; the ratio of arterial lactate to
pyruvate concentrations provides a clearer manifesta­
tion of the threshold behaviour,
3) Logarithmic arterial lactate and bicarbonate concen­
tration functions. This approach transforms the arterial
lactate concentration (or arterial standard bicarbonate
concentration, standard ([HC03\])) response to a loga­
rithmic function. This provides an even clearer discri­
mination of the lower (i.e, subthreshold) and higher (i.e.
suprathreshold) regions of the response. It typically lin­
earizes the more sharply rising phase of the response,
intersecting the lower phase of the response, which com­
monly displays a small positive slope. It is important,
however, that the appropriate sampling site is used for
determining LT. Arterial (or mixed venous, in some
instances) blood is most appropriate. Properly "arterialized" venous blood is acceptable.
Noninvasive estimators of LT (ventilatory equivalent
method and V-slope method) [194, 195] are described
below. As indicated previously (see section on Equip­
ment and quality control programme), the discriminability
of LT can be enhanced in some cases by utilizing a par­
ticular cluster of ventilatory and pulmonary gas exchange
responses.
Ventilatory equivalent method. The compensatory hy­
perventilation for the metabolic acidosis of heavy and
severe exercise typically coincides with the increase in
arterial lactate concentration and the decrease in arter­
ial standard [HC03~]. As a test for this, the work rate
is incremented in a quasi-steady state manner (i.e. incre­
ment duration of 3-4 min or more); V'E begins to increase
more rapidly than V'C02 and V o 2 (i.e. the ventilatory
equivalents for oxygen and carbon dioxide (Ve/Vo2, V eI
V co2) both increase and P&,C02 is, therefore, reduced.
In contrast, however, when the work rate incremen­
tation rate is more rapid, the compensatory hyperventi­
lation is strikingly attenuated. In this situation, there is
a range of work rates immediately above LT within
which V'E and V'co2increase in approximately the same
proportionality as for moderate exercise; i.e. Ve/Vco2
does not increase. Under these conditions, V co 2 has
contributions both from metabolic sources and from bi­
carbonate buffering reactions. Therefore, as Pa,c02does
not fall in this region, it has been termed the range of
1,isocapnic buffering". However, the additional Ve that
clears this augmented load of carbon dioxide is, there­
fore, out of proportion to V'o2 causing V,e/Vo2 to in­
crease. Respiratory compensation for the lactic acidosis
CLINICAL EXERCISE TESTING
2683
(i.e. with Pa,CO2 actually decreasing) only begins for
rapid incremental tests at a work rate which is typica­
lly about midway between LT and V'02,max. This has
been termed the "respiratory compensation point". It is,
therefore, important to recognize that the respiratory
compensation point is a function of the incrementation
rate of the test, whereas LT is not.
The phenomenon of isocapnic buffering consequent­
ly helps rule out nonspecific hyperventilation (such as
is seen in some excitable subjects and also in patients
with McArdle's syndrome) as a "false positive" lactate
threshold. For an appropriate rapid incremental exer­
cise protocol, LT has typically been taken as the V'o2
(not the work rate) at which the alveolar (end-tidal) Po2
(Pet,0 2) and the ventilatory equivalent for oxygen (V'E/
V o2) start to rise systemically without a simultaneous
fall in P et,co 2.
Some individuals (such as those with intensive periph­
eral chemoreceptors or those with appreciably increa­
sed airways resistance), may have a compromised ability
to generate the required increases in V'E at higher work
rates. As a result, if there is no, or possibly little, addition­
al ventilatory response, it proves difficult, or impossible,
to discriminate LT according to these ventilatory-related
criteria. The V-slope approach was developed to over­
come this challenge to discrimination.
Sj and S2 regions are perceived not to be sufficiently
linear for the intersection to be meaningful, it has been
proposed that the point on the curve at which a line of
slope 1.0 is tangential will provide an adequate approx­
imation of the threshold [196].
The V-slope method. With this procedure, LT is iden­
tified from the relationship between V'C02 and V'0 2. The
acceleration of the rate of increase in V'C02 relative to
V'02 above the LT provides evidence for an increased,
but still essentially linear, slope (S2, fig- 8), within the
isocapnic buffering phase. The intersection of the two
linear phases (Sj and S2, fig. 8) has been shown to agree
closely with the beginning of the increase in arterial lac­
tate concentration and arterial lactate/pyruvate concen­
tration ratio and even more closely with the decrease
in arterial standard [C03H-], i.e. a small amount of the
initial increase in muscle lactate is buffered by nonbi­
carbonate buffers [2, 147], and hence does not yield
increased V'C0 2. The consequence for LT discrimina­
tion, however, is likely to be disappearingly small dur­
ing incremental testing. On those occasions, when the
It, thus, represents the potential for further increase in
ventilation during maximal (or peak) exercise.
4-
3e
E
CM
2-
O
O
1-
0
0
1
2
3
in “1
V'o2 L-min
Fig. 8. - Estimation of lactate threshold (LT) using the V-slope
method (see text for further explanations). The intersection of the
slopes of the two linear phases (Si and S2) corresponds to the LT,
For definitions, see legend to table 5.
Ventilation, Minute ventilation L-min-1. Volume of gas
expired (V'E) or inspired (V'l) in one minute, expressed
in BTPS conditions (see below).
Tidal volume ( Vr, mL or L). Volume of gas inspired
(or expired) during each breathing cycle. It is comput­
ed as the ratio of minute ventilation (L'mhr1) to respi­
ratory frequency (breaths-mur1).
Respiratory frequency (/R, in breaths-min”1). Number of
breathing cycles per minute. It is computed as the ratio
60 divided by the total time of the breathing cycle (s).
Ventilatory reserve (VR, dimensionless). Difference bet­
ween ventilatory capacity during maximum exercise (es­
timated as maximum voluntary ventilation (M W )) and
minute ventilation at peak exercise, expressed as a per­
centage of M W . It is computed as:
VR % = ((M W - peak V'e)/MVV>100
Respiratory blood gases. Arterial Po2 (Pa,0 2, mmHg).
Oxygen partial pressure in arterial blood.
Arterial PC02 (Pa,C02 mmHg), Carbon dioxide partial
pressure in arterial blood.
Alveolar-arterial oxygen difference (DA-a,o2, mmHg).
Difference between ideal alveolar partial pressure of
oxygen (Pa,o2) and measured Pa,o2. The DA-a,o2is com­
puted using the alveolar gas equation described in the
opening section:
DA-a,02 = (Pl,02 - Pa,C0 2/RER +
(Pa,C02-Pr,0 2*(1-RER)/RER) - Pa,02
where RER is the respiratory exchange ratio (RER=
V'C02/V'02). This equation is derived from mass balance
considerations of oxygen and carbon dioxide exchange
between the air and the alveolar gas and assumes con­
stancy of lung nitrogen stores. Thus, even if oxygen,
carbon dioxide and V'E are changing, the DA-a,02 can
be validly computed if there is nitrogen balance between
inspiration and expiration. It is important to recognize
that the absolute blood gas values and the DA-a,o2 yield
different, but complementary information. The DA-a,02
primarily reflects pulmonary defects in gas exchange
caused by V'a/Q' mismatching, diffusion limitation and
shunt, either alone or in combination. It can, however,
be modified by changes in cardiac output or ventilation
even in the absence of change in V'a/Q' distribution.
Dead space to tidal volume ratio (Vd/Vt dimensionless). Portion of tidal volume (Vt ) ventilating physio­
logic dead space (VD). It is an index of the efficiency
2684
J. ROCA, B.J. WHIPP
of the lung as carbon dioxide exchanger. It is compu­
ted as the ratio of the difference between Pa,C02 and
mixed expired Pco 2 (Pa,C02- Pe,C0 2) to Pa,co2:
References
1.
VD/VT % = ((jPafC02 - PE,CO2)/Pa,CO2>100
Cardiovascular. Cardiac frequency (/C, beats’miir1). Num­
ber of cardiac beats per minute. It is computed as the
ratio of number of complete beats during 1 min to the
total time of the cardiac cycle.
Heart rate reserve (HRR, dimensionless). Ratio of the
difference between the predicted highest cardiac fre­
quency at maximum exercise and the measured peak
cardiac frequency to the predicted highest cardiac fre­
quency at maximum exercise, expressed as a percent­
age.
2.
3.
4.
HRR = ((Predicted maximum fc - peak fc)/
predicted maximum /C)* 100
5.
0 2 pulse (V'Ojft:, mL). Corresponds to the amount of
oxygen uptake per cardiac cycle. The 0 2 pulse is num­
erically equal to the product of the stroke volume and
the arterial-to-venous oxygen content difference. Con­
sequently, inferences which are drawn about one of the
contributory variables are highly dependent on assump­
tions regarding the response of the other(s).
6.
Acid-base status. Arterial pH (pHa). Inverse of the dec­
imal logarithm of the arterial blood hydrogen ion con­
centration ([H+]), It indicates the global acid-base status
(respiratory and metabolic components) of the arterial
blood.
Pa,C02 (described above). Reflects the contribution of
7.
8.
9.
10.
11.
the respiratory component to the acid-base status .
Base excess (BE, mEq-L1)- Indicates the contribution
of the metabolic component (nonvolatile substances) to
the acid-base status of the blood. The BE is computed
using the Henderson-Hasselbach equation with mea­
sured values of pHa and Pa,C0 2. Haemoglobin concen­
tration and oxyhaemoglobin saturation influence the BE
calculation by modulating the buffer capacity of the
blood. The information provided by this variable is sim­
ilar to that obtained estimating the standard [~C03H].
Measurement conditions. STPD. Gas volume measure­
ments at standard conditions of temperature (0 °C) and
pressure (760 mmHg (101.3 kPa)), free of water vapour
(partial pressure of water vapour Ph2o = 0 kPa).
BTPS. Gas volume measurements at body temperature
(37DC), ambient atmospheric pressure and saturated with
water vapour at body temperature (Ph2o = 47 mmHg
(6.2 kPa)).
12.
13.
14.
15.
16.
17.
18.
Acknowledgements: The members of the Task
Force on Clinical Exercise Testing are grateful to M.
Hillenius and P. Chapman for their skilful assistance
in the preparation of the manuscript and to F. Burgos
for preparing the figures.
19.
Weisman IM , Zeballos R l. An integrated approach to
the interpretation of CPET. In : Weisman IM, Zeballos
R J, eds. Clinical Exercise Testing; Clinics in Chest
Medicine. Philadelphia, W.B. Saunders, 1994.
Wasserman K, Hansen JE, Sue D Y , Whipp BJ, Casaburi
R. Principles of Exercise Testing and Interpretation.
Philadelphia, Lea & Febiger, 1994.
Killian KJ, Leblanc P, Martin DH, Summers E, Jones
NL, Campbell EJM. Exercise capacity and ventilatory,
circulatory, and symptom limitation in patients with
chronic airflow limitation. Am Rev Respir Dis 1992;
146: 935-940.
Ortega F, Montemayor T, Sanchez A, Cabello F, Castillo
J. Role of cardiopulmonary exercise testing and the cri­
teria used to determine disability in patients with severe
chronic obstructive pulmonary disease. Am J Respir
Crit Care Med 1994; 150: 747-751.
Weisman IM , Zeballos RJ. Cardiopulmonary exercise
testing. Pulmonary Critical Care Update series 1995;
11: 1-9.
Oren A, Sue D Y , Hansen JE, Torrance DJ, Wasserman
K. The role of exercise testing in impairment evalua­
tion. Am Rev Respir Dis 1987; 135: 230-235.
Cotes JE, Zejda J, King B. Lung function impairment
as a guide to exercise limitation in work-related lung
disorders. Am Rev Respir Dis 1988; 137: 1089-1093.
Sue DY. Exercise testing in the evaluation of impair­
ment and disability. Clin Chest Med 1994; 15: 369-387.
Curtis JR, Deyo RA, Hudson LD. Health-related qual­
ity of life among patients with chronic obstructive pul­
monary disease. Thorax 1994; 49: 162-170.
Knight DR, Schaffartzik W , Poole DC, Hogan M C,
Bebout DE, Wagner PD. Effects of hyperoxia on max­
imal leg 0 2 supply and utilization in men. J Appi Phy­
siol 1993; 75: 2586-2594.
Belman MJ. Exercise in chronic obstructive pulmon­
ary disease. In: Ryan A J, Allman FLJ, eds. Sports
Medicine. New York, Academic Press, 1989; pp. 647669.
Hansen JE, Sue D Y , Wasserman K. Predicted values
for clinical exercise testing. Am Rev Respir Dis 1984;
129: S49-S50.
Nery LE, Wasserman K, Andrews JD, Huntsman DJ,
Hansen JE, Whipp BJ. Ventilatory and gas exchange
kinetics during exercise in chronic airways obstruction.
J Appi Physiol 1982; 53: 1594-1602.
Linnarsson D. Dynamics of pulmonary gas exchange
and heart rate at start and end of exercise. Acta Physiol
Scand 1974; 415: 1-68.
Whipp BJ. Dynamics of pulmonary gas exchange dur­
ing exercise in man. Circulation 1987; 76 (6 part 2) VI:
18-28.
Dantzker DR, D 'Alonzo GE. The effect of exercise on
pulmonary gas exchange in patients with severe chron­
ic obstructive pulmonary disease. Am Rev Respir Dis
1986; 134: 1135-1139.
Gallagher CG. Exercise limitation and clinical exercise
testing in chronic obstructive pulmonary disease. Clin
Chest Med 1994; 15: 305-326.
Agusti AG, Barberà JA, Roca J, Rodriguez-Roisin R,
Wagner PD, Agusti-Vidal A. Hypoxic pulmonary vaso­
constriction and gas exchange during exercise in chro­
nic obstructive pulmonary disease. Chest 1990; 97:
268-275.
Jones NL, K illian KJ. Exercise in chronic airway
CLINICAL EXERCISE TESTING
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
obstruction. In: Bouchard C, ed. Exercise Fitness and
Health. Champaign, IL, USA. Human Kinetics Publish­
ers, 1990; pp. 547-559.
Hansen JE. Respiratory abnormalities: exercise evalu­
ation of the dyspnoeic patient. In: Leff AR, ed. Cardio­
pulmonary Exercise Testing. New York, Grune and
Stratton, 1986; pp. 68-88.
Wagner PD, Gale GE. Ventilation-perfusion relation­
ships. In: Whipp BJ, Wasserman K, eds. Pulmonary Phy­
siology and Pathophysiology of Exercise. New York,
Dekker, 1991; pp. 121-142.
D'Alonzo GE, Gianotti LA , Pohil RL. Comparison of
progressive exercise performance of normal subjects and
patients with primary pulmonary hypertension. Chest
1987; 92: 57-62.
Agusti AG, Roca J, Rodriguez-Roisin R, Gea J, Xaubet
A, Wagner PD. Mechanisms of gas exchange impair­
ment in idiopathic pulmonary fibrosis. Am Rev Respir
Dis 1991; 143: 219-225.
Dantzker DR, D 1Alonzo GE, Bower JS, Pop at K, Crevey
BJ. Pulmonary gas exchange during exercise in patients
with chronic obliterative pulmonary hypertension. Am
Rev Respir Dis 1984; 130: 412-416.
Jones NL, Rebuck AS. Tidal volume during exercise in
patients with diffuse fibrosing alveolitis. Bull Eur
Physiopathol Respir 1979; 15: 321-327.
Potter W A, Olafsson S, Hyat RE. Ventilatory mechan­
ics and expiratory flow limitation during exercise in
patients with obstructive lung disease. J Clin Invest
1971; 50: 910-919.
Gallagher C . Exercise and chronic obstructive pulmon­
ary disease. Med Clin N Am 1990; 74: 619-641.
Johnson BD, Dempsey JA. Demand versus capacity in
the aging pulmonary system. Ex Sports Sci Rev 1991;
19: 171-210.
Younes M. Determinants of thoracic excursions during
exercise. In: Whipp BJ, Wasserman K, eds. Pulmonary
Physiology and Pathophysiology of Exercise. New York,
Dekker, 1991; pp. 1-65.
Marciniuk DD, Gallagher CG. Clinical exercise testing
in interstitial lung disease. Clin Chest Med 1994; 15:
287-303.
Scano G , Gigliotti F, van Meerhaeghe A, et al. Influence
of exercise and C 0 2 on breathing pattern in patients
with chronic obstructive lung disease (COLD). Eur
Respir J 1988; 1: 139-144.
Raimondi AC, Edwards RHT, Denison DM, et al. Ex­
ercise tolerance breathing a low density gas mixture,
35% oxygen and air in patients with chronic obstruc­
tive bronchitis. Clin Sci 1970; 39: 675-685.
Nattie EE, Tenney SM. The ventilatory response to
resistance unloading during muscular exercise. Respir
Physiol 1970; 20: 249-262.
Hussain SNA, Pardy R L, Dempsey JA. Mechanical
impedance as determinant of inspiratory neural drive
during exercise in humans. J Appl Physiol 1985; 59:
365-375.
Wagner PD, Dantzker D R , Dueck R, et al. Ventilationperfusion inequality in chronic obstructive pulmonary
disease. J Clin Invest 1917; 59: 203-216.
Stewart RI, Lewis CM. Cardiac output during exercise
in patients with COPD. Chest 1986; 89: 199-205.
Wagner PD, Hoppeler H, Saltin B. Determinants of
maximal oxygen uptake. In: Crystal RG, West JB, eds.
The Lung: Scientific Foundations. New York, Raven
Press, 1991; pp. 1585-1594.
Wahren J, Felig P, Havel RJ, Jorfeldt L, Pemow B,
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
2685
Saltin B. Amino acid metabolism in McArdle's syn­
drome. N Engl J Med 1973; 288: 774-777.
Martinez FJ, Stanopoulos I , Acero R , B ecker FS, Pickering
R, Beamis JF. Graded comprehensive cardiopulmonary
exercise testing in the evaluation of dyspnea unexplain­
ed by routine evaluation. Chest 1994; 105: 168-174.
Weisman IM, Zeballos RJ. Clinical evaluation of un­
explained dyspnea. Cardiologia 1996; 41: 621-634.
Keogh BA, Lakatos E, Price D, et al. Importance of the
lower respiratory tract in oxygen transfer: exercise test­
ing in patients with interstitial and destructive lung
disease. Am Rev Respir Dis 1984; 129: S76-S80.
Kelley MA, Daniele RP. Exercise testing in interstitial
lung disease, Clin Chest Med 1984; 5: 145-156.
Fulmer JD, Roberts WC, Von Gal ER, Crystal RG. Mor­
phologic-physiologic correlates of the severity of fibro­
sis and degree of cellularity in idiopathic pulmonary
fibrosis. J Clin Invest 1979; 63: 665-676.
Agusti C, Xaubet A, Roca J, Agusti AG, RodriguezRoisin R. Interstitial pulmonary fibrosis with and with­
out associated collagen vascular disease: results of a 2
years follow up. Thorax 1992; 47: 1035-1040.
Nickoladze GD. Functional results of surgery for bul­
lous emphysema, Chest 1992; 101: 119-122.
Cooper JD,Trulock EP, Triantafillou AN, et al. Bilateral
pneumectomy (volume reduction) for chronic obstruc­
tive pulmonary disease. J Thorac Cardiovasc Surg 1995;
109: 106-119.
D'Alonzo GE, Gianotti L, Dantzker DR. Noninvasive
assessment of haemodynamic improvement during chro­
nic vasodilator therapy in obliterative pulmonary hy­
pertension. Am Rev Respir Dis 1986; 133: 380-384.
Biggar DG, Malen JF, Trulock EP, Cooper JD. Pulmon­
ary rehabilitation before and after lung transplantation.
In: Casaburi R, Petty TL, eds. Principles and Practice
of Pulmonary Rehabilitation. Philadelphia, W.B. Saun­
ders Co., 1993; pp, 459-467.
Rubin LJ. ACCP Consensus Statement. Primary pul­
monary hypertension. Chest 1993; 104: 236-250.
Rhodes J, Barst RJ, Garofano RP, Thoele DG, Gersony
WM. Haemodynamic correlates of exercise function in
patients with primary pulmonary hypertension. J Am
Coll Cardiol 1991; 18: 1738-1744.
Nixon PA, Orenstein D M , Kelsey SF, Doershuk C F . The
prognostic value of exercise testing in patients with
cystic fibrosis. N Engl J Med 1992; 327: 1785-1788.
Older P, Hall A. The role of cardiopulmonary exercise
testing for preoperative evaluation of the elderly. In:
Wasserman K, ed. Exercise Gas Exchange in Heart
Disease. New York, Armonk: Futura Publishing Com­
pany, 1996; Chapter 20.
Shoemaker WC, Appel PL, Kram HB. Role of oxygen
debt in the development of organ failure sepsis, and
death in high-risk surgical patients. Chest 1992; 102:
208-215.
Marshall MC, Olsen GN. The physiologic evaluation
of the lung resection candidate. Clin Chest Med 1993;
14: 305-320.
Ferguson MK, Little L, Rizzo L, et a l Diffusing capa­
city predicts morbidity and mortality after pulmonary
resection. J Thorac Cardiovasc Surg 1988; 96: 894-900.
Kearney DJ, Lee TH, Reilly JJ, DeCamp MM, Sugarbaker
DJ. Assessment of operative risk in patients undergo­
ing lung resection: importance of predicted pulmonary
function. Chest 1994; 105: 753-759.
Markos J, Mullan BP, Hillman DR, et aL Preoperative
assessment as a predictor of mortality and morbidity
2686
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
J. ROC A, B J . WHIPP
after lung resection. Am Rev Respir Dis 1989; 139:
902-910.
Olsen GN. The evolving role of exercise testing prior
to lung resection. Chest 1989; 95: 218-225.
Boysen PS. Perioperative management of the thoraco­
tomy patient. Clin Chest Med 1993; 14: 321-333.
Bolliger CT, Wyser C, Roser H, Solfcr M , Perruchoud
AP. Lung scanning and exercise testing for the predic­
tion of postoperative performance in lung resection can­
didates at increased risk for complications. Chest 1995;
108: 341-348.
American Thoracic Society: ATS Statement. Standards
for the diagnosis and care of patients with chronic ob­
structive pulmonary disease. Am J Respir Crit Care Med
1995; 152: 77-120.
Sciurba FC, Rogers R M , Keenan RJ, et al. Improve­
ment in pulmonary function and elastic recoil after lungreduction surgery for diffuse empysema. N Engl J Med
1996; 334: 1095-1099.
American Thoracic Society. Pulmonary rehabilitation.
Am Rev Respir Dis 1981; 124: 663-666.
Fishman AP. Pulmonary rehabilitation research. Am J
Crit Care 1994; 149: 825-833.
Make BJ. Pulmonary rehabilitation: myth or reality.
Clin Chest Med 1986; 7: 519-540.
Casaburi R , Petty TL. Principles and Practice of Pulmon­
ary Rehabilitation. Philadelphia, W .B. Saunders, 1993.
McGavin C R, Gupta SP, Lloyd EL, et al. Physical reha­
bilitation for the chronic bronchitic: results of a con­
trolled trial of exercises in the home. Thorax 1977; 32:
307-311.
Celli B, Rassulo J, Make BJ. Dyssynchronous breath­
ing is associated with arm but not leg exercise in chro­
nic obstructive pulmonary disease. N Engl J Med 1986;
314: 1485-1490.
Guyatt GH , Townsend M , Berman LB, Pugsely SD.
Quality of life in patients with chronic airflow limita­
tion. Br J Dis Chest 1987; 81: 45-54.
Atkins CJ, Kaplan RM ,T im m s RM , Reinsch S , Lofback
K. Behavioral exercise programs in the management of
chronic obstructive pulmonary disease. J Consult Clin
Physiol 1984; 52: 591-603.
Ries AL, Kaplan R M , Limberg TM, Prewitt LM . Effects
of pulmonary rehabilitation on physiologic and psy­
chosocial outcomes in patients with chronic obstructive
pulmonary disease. Ann Intern Med 1995; 122: 823832.
Couser JIJ, Guthmann R , Hamadeh M A , Kane CS. Pul­
monary rehabilitation improves exercise capacity in
older elderly patients with COPD. Chest 1995; 107:
730-734.
Casaburi R , Patessio A , loli F, Zanaboni S, Donner CF,
Wasserman K. Reductions in exercise lactic acidosis
and ventilation as a result of exercise training in pati­
ents with obstructive lung disease, Am Rev Respir Dis
1991; 143: 9-18.
Griffiths T, Gregory SE, Ward SA, Saunders KB, Whipp
BJ. Effect of domiciliary exercise training on ventila­
tory and gas-exchange kinetics in patients with severe
COPD. Eur Respir J 1995; 8: 502s.
ATS Statement. Evaluation of impairment/disability
secondary to respiratory disorders. Am Rev Respir Dis
1986; 133: 1205-1209.
World Health Organization. International classification
of impairments, disabilities, and handicaps. A manual
of classification relating to the consequences of disease.
Published for trial purposes in accordance with resolu­
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
tion WH029.35 of the Twenty-ninth World Health
Assembly, May 1976. Geneva, Schuler, A.G. 1980.
Marshall SE, Kramer M R, Lewiston NJ, Starnes VA,
Theodore J. Selection and evaluation of recipients for
heart-lung and lung transplantation. Chest 1990; 98:
1488-1494.
Mudge GH, Goldstein S, Addonizio LJ, et al. 24th
Bethesda Conference: Cardiac Transplantation. Task
force 3: recipients guidelines/prioritization. J Am Coll
Cardiol 1993; 22: 21-31,
Warner Stevenson L. Role of exercise testing in the
evaluation of candidates for cardiac transplantation. In:
Wasserman K, ed. Exercise Gas Exchange in Heart
Disease. New York, Armonk, Futura Publishing Com­
pany, 1996; Chapter 19.
Craven JL, Bright J, Dear CL. Psychiatric, psychoso­
cial, and rehabilitative aspects of lung transplantation.
Clin Chest Med 1990; 11: 247-257.
Casaburi R. Deconditioning. In: Fishman AP, ed. Pul­
monary Rehabilitation. New York, Marcel Dekker, 1996;
pp. 213-230.
The Toronto Lung Transplant Group. Experience with
single-lung transplantation for pulmonary fibrosis. JAMA
1988; 259: 2258-2262.
Howard DK, Iademarco EJ, Trulock EP. The role of
cardiopulmonary exercise testing in lung and heart-lung
transplantation. Clin Chest Med 1994; 15: 405-420.
Levy RD, Ernst P, Levine SM, et al. Exercise perfor­
mance after lung transplantation. J Heart Lung Trans­
plant 1993; 12: 27-33.
Roca J, Whipp BJ, eds. Clinical Exercise Testing. Eur
Respir Mon 1997; (in press).
Kavanagh T, Yacoub MH, Mertens D J, Kennedy J,
Campbell RB, Sawyer P. Cardiorespiratory responses
to exercise training after orthotopic cardiac transplan­
tation. Circulation 1988; 77: 162-171.
Jones NL. Clinical exercise testing. Philadelphia, W.B.
Saunders, 1980.
Whipp BJ, Davis JA, Torres F, Wasserman K. A test
to determine parameters of aerobic function during ex­
ercise. J Appl Physiol 1981; 50: 217-221.
Sterk PJ, Fabbri LM , Quanjer PH, et al. Airway res­
ponsiveness. Standardized challenge testing with phar­
macological physical and sensitizing stimuli in adults.
Eur Respir J 1993; 6: 53-83.
Balke B, Ware RW, An experimental study of "physi­
cal fitness" of Air Force personnel. US Armed Forces
Med J 1959; (Abstract).
Zeballos RJ, Weisman IM . Behind the scenes of car­
diopulmonary exercise testing. Clin Chest Med 1994;
15: 193-213.
Mador MJ, Rodis A, Magalang UJ. Reproducibility of
borg scale measurements of dyspnea during exercise in
patients with COPD. Chest 1995; 107: 1590-1597.
Hansen JE, Participant responses to blood gas profi­
ciency testing reports. Chest 1992; 101: 1240-1244.
Clark JS, Votteri B, Arriagno RL, et al. Noninvasive
assessment of blood gases. Am Rev Respir Dis 1992;
145: 220-232.
Hughes JMB. Blood gas estimations from arterialized
capillary blood versus arterial puncture: are they dif­
ferent? Eur Respir J 1996; 9: 184-185.
Sauty A, Uldry C, Deb6taz L, Leuenberger P, Fitting
J. Differences in Po2 and Po2between arterial and arte­
rialized earlobe samples. Eur Respir J 1996; 9: 186-189.
Hughson RL, Weisiger KH, Swanson GD. Blood lac­
tate concentration increases as a continuous function in
CLINICAL EXERCISE TESTING
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
progressive exercise. J App I Physiol 1987; 62: 19751981.
Anderson SD. Exercise-induced asthma. In: O'Byrne P,
Thompson NC, eds. Manual of Asthma Management.
London, W.B. Saunders Co., 1995; pp. 621-643.
Cropp GJA. The exercise bronchoprovocation test: stan­
dardization of procedures and evaluation of response. J
Allergy Clin Immunol 1979; 64: 627-633.
Haby MM, Peat JK, Mellis CM, Anderson SD, Woolcock
AJ, An exercise challenge for epidemiological studies
of childhood asthma: validity and repeatability. Eur
Respir J 1995; 8: 729-736.
Strauss RE, McFadden ER, Ingram RH, Deal EC, Jaeger
JJ. Influence of heat and humidity on the airway ob­
struction induced by exercise in asthma. J Clin Invest
1978; 61: 433-440.
McFadden ER, Lenner K A M , Kingman PS. Post exer­
tional airway re warming and thermally induced asth­
ma: new insights and pathophysiology and possible
pathogenesis. J Clin Invest 1986; 78: 18-25.
Anderson SD, Schoeffel RE, Follet R, Perry CP, Daviskas
E, Kendall M. Sensitivity to heat and water loss at rest
and during exercise in asthmatic patients. Eur J Respir
Dis 1982; 63: 459-471.
Anderson SD, Daviskas E. The airway microvasculature and exercise induced asthma. Thorax 1992; 47:
748-752.
Anderson SD, Silverman M , König F, Godfrey S. Ex­
ercise-induced asthma. Brit J Dis Chest 1975; 69: 1-39.
Givoni B, Goldman RF. Predicting metabolic energy
cost. J Appl Physiol 1971; 30: 429-433.
Silverman M , Anderson SD, Metabolic cost of tread­
mill exercise in children. J Appl Physiol 1972; 33: 696698.
Noviski N, Bar-Yishay E, Gur I, Godfrey S, Exercise
intensity determines and climatic conditions modify
the severity of exercise-induced asthma. Ant Rev Respir
Dis 1987; 136: 592-594.
Phillips Y Y , Jaeger JJ, Laube BL, Rosenthal RR. Eucapnic
voluntary hyperventilation of compressed gas mixture:
a simple system for voluntary hyperventilation of com­
pressed gas mixture: a simple system for bronchial chal­
lenge by respiratory heat loss. Am Rev Respir Dis 1985;
131: 31-35.
Rod well LT, Anderson SD, Du Toit J, Seale JP. Differ­
ent effects of inhaled amiloride and furosemide on air­
way responsiveness to dry air challenge in asthmatic
subjects. Eur Respir J 1993; 6: 855-861.
Anderson SD, Rodwell LT, Du Toit J, Young IH.
Duration of protection of inhaled salmeterol in exer­
cise-induced asthma. Chest 1991; 100: 1254-1260.
Woolley M , Anderson SD, Quigley B. Duration of pro­
tective effect of terbutaline sulphate and cromolyn sodi­
um alone and in combination on exercise-induced asthma.
Chest 1990; 97: 39-45,
Kemp JP, Dockhom R J, Busse W W , Blaecker NR, Van
As A. Prolonged effect of inhaled salmeterol against
exercise-induced bronchospasm. Am J Respir Crit Care
Med 1994; 150: 1612-1615.
Bruce RA. Exercise testing of patients with coronary
artery disease. Ann Clin Res 1971; 3: 323-332.
Patterson JA, Naughton J, Pietras RJ, Gumar RN.
Treadmill exercise in assessment of patients with car­
diac disease. Am J Cardiol 1972; 30: 757-762.
Cotes JE. Lung function: Assessment and Application
in Medicine. Fifth edition. London, Blackwell Scien­
tific Publishers 1993.
2687
117. Buchfuhrer MJ, Hansen JE, Robinson TE, Sue DY,
Wasserman K, Whipp BJ. Optimizing the exercise proto­
col for cardiopulmonary assessment (Abstract). J Appl
Physiol 1983; 55: 1558-1564.
118. Astrand PO, Rodahl K. Textbook of Work Physiology.
New York, McGraw-Hill, 1986.
119. Hermansen L, Saltin B. Oxygen uptake during maxi­
mal treadmill and bicycle exercise. J Appl Physiol 1969;
26: 1-31.
120. McArdle WD, Katch RI, Pechar GS. Comparison of
continuous and discontinuous treadmill and bicycle tests
for max V o 2. Med Sci Sports Exerc 1973; 5; 156-160.
121. McKay GA, Banister EW. A comparison of maximum
oxygen uptake determination of bicycle ergometry at
various pedaling frequencies and by treadmill running
at various speeds. Eur J Appl Physiol 1976; 35: 191—
200.
122. Koyal SN, Whipp BJ, Huntsman D , Bray GA, Wasserman
K. Ventilatory responses to the metabolic acidosis of
treadmill and cycle ergometry. J Appl Physiol 1976; 40:
864-867.
123. Whipp BJ. The Physiological and Energetic Basis of
Work Efficiency. Bray GA, ed. Obesity in Perspective.
Washington DC, U.S. Govt. Printing Office, 1976; pp.
121-126.
124. Whipp BJ, Bray GA, Koyal SN, Wasserman K. Exercise
Energetics and Respiratory Control in Man Following
Acute and Chronic Elevation of Calorie Intake, Obe­
sity in Perspective. Bray GA, ed. Washington DC, U.S.
Govt. Printing Office, 1976; pp. 157-163,
125. Wasserman K, Whipp BJ. Exercise physiology in hea­
lth and disease (state of the Ait). Am Rev Respir Dis
1975; 112: 219-249.
126. Lanooy C, Bonjer FH. A hyperbolic ergometer for
cycling and cranking. J Appl Physiol 1956; 9: 499-500.
127. American Heart Association Subcommittee on Rehabil­
itation Target Activity Group, Hellerstein HKC, Specifica­
tion for exercise testing equipment. Circulation 1979;
59: 849A-854A.
128. Davis JA, Whipp BJ, Lamarra N, Huntsman DJ, Frank
MH, Wasserman K. Effect of ramp slope on measure­
ment of aerobic parameters from the ramp exercise test.
Med Sci Sports Exerc 1982; 14: 339-343.
129. Hansen JE, Casaburi R, Cooper DM, Wasserman K.
Oxygen uptake as related to work rate increment dur­
ing cycle ergometer exercise. Eur J Appl Physiol 1988;
57: 140-145.
130. Zhang YY, Johnson MC, Chow N, Wasserman K. Effect
of exercise testing protocol on parameters of aerobic
function. Med Sci Sports Exerc 1991; 23: 625-630.
131. Myers J, Buchanan N, Walsh D, et at. Comparison of
the ramp versus standard exercise protocols. J Am Coll
Cardiol 1991; 17: 1334-1342.
132. Tanner CS, Heise CT, Barber G. Correlation of the phy­
siologic parameters of a continuous ramp versus an
incremental James exercise protocol in normal child­
ren. Am J Cardiol 1991; 67; 309-312.
133. Naughton J. Exercise Testing. Physiological, Biochemi­
cal and Clinical Principles. New York, Futura Publishing Company, 1988.
134. Beaver WL. Water vapour corrections in oxygen con­
sumption calculations. J Appl Physiol 1973; 35: 928931.
135. Poole GW , Maskell RC. Validation of continuous
determination of respired gases during steady-state exer­
cise. J Appl Physiol 1975; 38: 736-738.
136. Spiro SG, Juniper E, Bowman P, Edwards RHT. An
2688
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
J. ROC A , B .J. WHIPP
increasing work rate test for assessing the physiologi­
cal strain of submaximal exercise. Clin Sci Mol Med
1974; 46: 191-206.
Beaver W L, Wasserman K , Whipp BJ. On-line com­
puter analysis and breath-by-breath graphical display of
exercise function tests. J Appl Physiol 1973; 34: 128132.
Noguchi H, Ogushi Y , Yoshiya N, Itakura N, Yamabayashi
H. Breath-by-breath Vco2 and Vo2 require compensa­
tion for transport delay and dynamic response. J Appl
Physiol 1982; 52: 79-84.
Yamamoto Y , Takei Y , Mokushi K, Morita H, Mutoh
Y , Miyashita M. Breath-by-breath measurement of alve­
olar gas exchange with a slow-response gas analyser.
Med Biol Eng Comput 1987; 25: 141-146.
B eaver W L , Lamarra N , Was serman K . B reath-by-breath
measurement of true alveolar gas exchange. J Appl
Physiol 1981; 51: 1662-1675.
Sue D Y , Hansen JE, Blias M , Wasserman K. Measure­
ment and analysis of gas exchange during exercise us­
ing a programmable calculator. J Appl Physiol 1980;
49: 456-461.
Lamarra N , Whipp BJ. Measurement of pulmonary gas
exchange. In: Maud PJ, Foster C, eds. Physiological
Assessment of Human Fitness. Champaign, IL, USA,
Human Kinetics, 1995; pp. 19-35.
Wilmore JH, Costill DL. Adequacy of the Haldane trans­
formation in the computation of exercise V'o2 in man.
J Appl Physiol 1973; 35: 85-98.
Swanson GD. Breath-by-breath considerations for gas
exchange kinetics. In: Ceretelli P, Whipp BJ, eds. Exercise
Bioenergetics and Gas Exchange. Amsterdam, Elsevier/
North Holland, 1980; pp. 211-222.
Dennis SC, Noakes TD, Bosch AN . Ventilation and
blood lactate increase exponentially during incremental
exercise. J Sports Sci 1992; 10: 437-449.
Wasserman K, Beaver W L, Whipp BJ. Mechanisms and
pattern of blood lactate increase during exercise in man.
Med Sci Sports Exerc 1986; 18: 344—352.
Beaver W L, Wasserman K , W hipp BJ. Bicarbonate
buffering of lactic acid generated during exercise. J Appl
Physiol 1986; 60: 472-478,
Clausen JL. Pulmonary Function Testing. Guidelines
and Controversies. Equipment, Methods and Normal
Values. Orlando, USA, Grune & Stratton» Inc. 1984;
p p .1-338.
Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF,
Peslin R , Yemault JC. Lung volumes and forced ven­
tilatory flows. Report Working Party: standardization
of lung function testing. Eur Respir J 1993; 6: 5-40.
'American Thoracic Society. Standardization of spiro­
metry: 1994 update. Am J Respir Crit Care Med 1995;
152: 1107-1136.
Finucane KE, Egan BA, Dawson SV. Linearity and fre­
quency response of pneumotachographs. J Appl Phy­
siol 1972; 32: 121-126.
Miller M R, Pincock AC. Linearity and temperature con­
trol of the Fleisch pneumotachograph. J Appl Physiol
1986; 60: 710-715.
Porszasz J, Barstow TJ, Wasserman L. Evaluation of a
symmetrically disposed Pitot tube flowmeter for mea­
suring gas flow during exercise. J Appl Physiol 1994;
77: 2659-2665.
Yoshiya I, Nakajima T, Nagai I, Jitsukawa S. A bidi­
rectional respiratory flowmeter using the hot-wire prin­
ciple. J Appl Physiol 1975; 38: 360-365.
Yoshiya I, Shimada Y , Tanaka K. Evaluation of a hot­
156.
157 .
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
wire respiratory flowmeter for clinical applicability. J
Appl Physiol 1979; 47: 1131-1135.
Yeh MP, Adams TD, Gardner RM , Yanowitz FG. Tur­
bine flow meter vs Fleisch pneumotachometer: a com­
parative study for exercise testing. J Appl Physiol 1987;
63: 1289-1295.
Cooper CB, Harris ND, Howard P. Evaluation of a tur­
bine flow meter (Ventilometer Mark 2) in the mea­
surement of ventilation. Respiration 1990; 57: 243-247.
Mogue L R , Rantala B . Capnometers. J Clin Monitoring
1988; 4: 115-121.
Fletcher GF, Froelicher VF, Hartley LH, Haskell WL,
Pollack ML. Exercise standards: a statement for health
professionals from the American Heart Association.
Circulation 1990; 82: 2286-2322.
American Heart Association. Exercise standards. Circula­
tion 1995; 91: 580-615.
Council on Scientific Affairs American Medical Asso­
ciation. The use of pulse oximetry during conscious
sedation. JAMA 1993; 270: 1463-1468.
Tobin MJ. Respiratory monitoring. JAMA 1990; 264:
244-251.
Ries AL, Farrow JT, Clausen JL. Accuracy of two ear
oximeters at rest and during exercise in pulmonary
patients. Am Rev Respir Dis 1985; 132: 685-689.
Carlin BW, Clausen JL, Ries AL. The use of cutaneous
oximetry in the prescription of long-term oxygen the­
rapy. Chest 1988; 94: 239-241.
Orenstein DM, Curtis SE, Nixon PA, Hartigan ER.
Accuracy of three pulse oximeters during exercise and
hypoxemia in patients with cystic fibrosis. Chest 1993;
104(4): 1187-1190.
Carlin BW , Clausen JL, Ries AL. The effects of exer­
cise testing on the prescription of oxygen therapy. Chest
1994; 106: 361-365.
Hansen JE, Casaburi R. Validity of ear oximetry in clin­
ical exercise testing. Chest 1987; 91: 333-337.
Ries AL, Prewitt LM, Johnson JJ. Skin color and ear
oximetry. Chest 1989; 96: 287-290.
Zeballos RJ, Weisman IM . Reliability of noninvasive
oximetry in black subjects during exercise and hypox­
ia. Am Rev Respir Dis 1991; 144: 1240-1244.
Ramsey M. Blood pressure monitoring: automated oscillometric devices. J Clin Monitoring 1991; 7: 56-57.
Robinson TR, Sue D Y, Huszczuk A, Weiler-Ravell D,
Hansen JE. Intra-arterial and cuff blood pressure respon­
ses during incremental cycle ergometry. Med Sci Sports
Exerc 1988; 20: 142-149.
Consolazio CF, Johnson RE, Pecora LJ. Physiological
measurements of metabolic functions in man. New
York, McGraw-Hill, 1963.
Huszczuk A, Whipp BJ, Wasserman K. A respiratory
gas exchange simulator for routine calibration in meta­
bolic studies. Eur Respir J 1990; 3: 465-468.
Clark JH, Greenleaf JE. Electronic bicycle ergometer:
a simple calibration procedure. J Appl Physiol 1971;
30; 440-442.
Van Praagh E, Bedu M, Roddier P, Coudert J. A sim­
ple calibration method for mechanically braked cycle
ergometers. Int J Sports Med 1992; 13: 27-30.
Russell JC, Dale JD. Dynamic torquemeter calibration of
bicycle ergometers. / Appl Physiol 1986; 61: 1217-1220.
Gardner RM, Schwartz R, Wong HC, et al. Percutaneous
indwelling radial-artery catheters for monitoring car­
diovascular function. N Engl J Med 1974; 290: 12271231.
Douard H, Mora B, Broustet JP. Epreuves d1effort et
CLINICAL EXERCISE TESTING
179.
180.
181.
182.
183.
184.
185.
186.
187.
tachycardies ventriculaires: l'expérience française. Arch
Mal Coeur 1987; 80: 263-270.
Simel DL, Samsa GP, Matcha DB. Likelihood ratios
with confidence: sample size estimation for diagnostic
test studies. J Clin Epidemiol 1991; 44: 763-770.
Evidence-Based Medicine Working Group. Evidencebased medicine: a new approach to the teaching of med­
icine. JAMA 1992; 268: 2420-2425.
Jones NL, Makrides L, Hitchcock C, Chypchar T,
McCartney N. Normal standards for an incremental pro­
gressive cycle ergometer test. Am Rev Respir Dis 1985;
131: 700-708,
Jones NL, Summers E, Killian KJ. Influence of age and
stature on exercise capacity during incremental cycle
ergometry in men and women. Am Rev Respir Dis 1989;
140: 1373-1380.
Blackie SP, Fairbam MS, McElvaney NG, Wilcox PG,
Morrison NJ, Pardy RL. Normal values and ranges for
ventilation and breathing pattern at maximal exercise.
Chest 1991; 100: 136-142.
Blackie SP, Fairbam MS, McElvaney NG, Morrison
NJ, Wilcox PG, Pardy RL. Prediction of maximal oxy­
gen update and power during cycle ergometry in sub­
jects older than 55 years of age. Am Rev Respir Dis
1989; 139: 1424-1429.
Fairbam MS, Blackie SP, McElvaney NG, Wiggs BR,
Paré PD, Pardy RL, Prediction of heart rate and oxy­
gen uptake during incremental and maximal exercise
in healthy adults. Chest 1994; 105: 1365-1369.
Bruce RA, Kusumi MS, Hosmer D. Maximal oxygen
intake and nomographic assessment of functional aer­
obic impairment in cardiovascular disease. Am Heart J
1973; 85: 546-562.
Vogel JA, Patton JF, Mello RP, Daniels WL. An ana­
188.
189.
190.
191.
192.
193.
194.
195.
196.
2689
lysis of aerobic capacity in a large United States pop­
ulation. J Appl Physiol 1986; 60: 494-500.
Pollock ML, Bohannon RL, Cooper KHE. A compar­
ative analysis of four protocols for maximal treadmill
stress testing. Am Heart J 1976; 92: 39-46.
Shephard RJ, Allen C, Benade AJ. The maximal oxy­
gen intake: an international reference standard of car­
diorespiratory fitness. Bull World Health Organization
1968; 38: 757-764.
Taylor HL, Buskirk E, Henschel A. Maximal oxygen
intake as an objective measure of cardio-respiratory per­
formance. J Appl Physiol 1955; 8: 73-80.
Wagner PD, Gale GE, Moon RE, Torre-Bueno JE, Stolp
BW, Saltzman HA. Pulmonary gas exchange in humans
exercising at sea level and simulated altitude. J Appl
Physiol 1986; 61: 260-270.
Young DS. Implementation of SI units for clinical lab­
oratory data. Style specifications and conversion tab­
les. Ann Intern Med 1987; 106: 114—129.
Cotes JE. SI units in respiratory medicine. Am Rev
Respir Med 1975; 112: 753-755.
Ward SA, Whipp BJ. Influence of body C 0 2 stores on
ventilatory-metabolic coupling during exercise. In:
Honda Y, Miyamoto K, Konno K, Widdicombe JG, eds.
Control of Breathing and its Modeling Perspective. New
York, Plenum Press, 1992; pp. 425-431.
Beaver W L, Wasserman K, Whipp BJ. A new method
for detecting the anaerobic threshold by gas exchange,
J Appl Physiol 1986; 60: 2020-2027.
Sue DY, Wasserman K, Moricca RB, Casaburi R. Meta­
bolic acidosis during exercise in patients with chronic
obstructive pulmonary disease. Use of v-slope method
for anaerobic threshold determination. Chest 1988; 94:
931-938.